bibtex_url
null | proceedings
stringlengths 42
42
| bibtext
stringlengths 197
848
| abstract
stringlengths 303
3.45k
| title
stringlengths 10
159
| authors
sequencelengths 1
34
⌀ | id
stringclasses 44
values | arxiv_id
stringlengths 0
10
| GitHub
sequencelengths 1
1
| paper_page
stringclasses 899
values | n_linked_authors
int64 -1
13
| upvotes
int64 -1
109
| num_comments
int64 -1
13
| n_authors
int64 -1
92
| Models
sequencelengths 0
100
| Datasets
sequencelengths 0
19
| Spaces
sequencelengths 0
100
| old_Models
sequencelengths 0
100
| old_Datasets
sequencelengths 0
19
| old_Spaces
sequencelengths 0
100
| paper_page_exists_pre_conf
int64 0
1
| type
stringclasses 2
values |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
null | https://openreview.net/forum?id=TbPv0qFnHO | @inproceedings{
leng2024beyond,
title={Beyond Euclidean: Dual-Space Representation Learning for Weakly Supervised Video Violence Detection},
author={Jiaxu Leng and Zhanjie Wu and Mingpi Tan and Yiran Liu and Ji Gan and Haosheng Chen and Xinbo Gao},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TbPv0qFnHO}
} | While numerous Video Violence Detection (VVD) methods have focused on representation learning in Euclidean space, they struggle to learn sufficiently discriminative features, leading to weaknesses in recognizing normal events that are visually similar to violent events (i.e., ambiguous violence). In contrast, hyperbolic representation learning, renowned for its ability to model hierarchical and complex relationships between events, has the potential to amplify the discrimination between visually similar events. Inspired by these, we develop a novel Dual-Space Representation Learning (DSRL) method for weakly supervised VVD to utilize the strength of both Euclidean and hyperbolic geometries, capturing the visual features of events while also exploring the intrinsic relations between events, thereby enhancing the discriminative capacity of the features. DSRL employs a novel information aggregation strategy to progressively learn event context in hyperbolic spaces, which selects aggregation nodes through layer-sensitive hyperbolic association degrees constrained by hyperbolic Dirichlet energy. Furthermore, DSRL attempts to break the cyber-balkanization of different spaces, utilizing cross-space attention to facilitate information interactions between Euclidean and hyperbolic space to capture better discriminative features for final violence detection. Comprehensive experiments demonstrate the effectiveness of our proposed DSRL. | Beyond Euclidean: Dual-Space Representation Learning for Weakly Supervised Video Violence Detection | [
"Jiaxu Leng",
"Zhanjie Wu",
"Mingpi Tan",
"Yiran Liu",
"Ji Gan",
"Haosheng Chen",
"Xinbo Gao"
] | NeurIPS.cc/2024/Conference | 2409.19252 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TZ5k9IYBBf | @inproceedings{
prabhu2024random,
title={Random Representations Outperform Online Continually Learned Representations},
author={Ameya Prabhu and Shiven Sinha and Ponnurangam Kumaraguru and Philip Torr and Ozan Sener and Puneet K. Dokania},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TZ5k9IYBBf}
} | Continual learning has primarily focused on the issue of catastrophic forgetting and the associated stability-plasticity tradeoffs. However, little attention has been paid to the efficacy of continually learned representations, as representations are learned alongside classifiers throughout the learning process. Our primary contribution is empirically demonstrating that existing online continually trained deep networks produce inferior representations compared to a simple pre-defined random transforms. Our approach embeds raw pixels using a fixed random transform, approximating an RBF-Kernel initialized before any data is seen. We then train a simple linear classifier on top without storing any exemplars, processing one sample at a time in an online continual learning setting. This method, called RanDumb, significantly outperforms state-of-the-art continually learned representations across all standard online continual learning benchmarks. Our study reveals the significant limitations of representation learning, particularly in low-exemplar and online continual learning scenarios. Extending our investigation to popular exemplar-free scenarios with pretrained models, we find that training only a linear classifier on top of pretrained representations surpasses most continual fine-tuning and prompt-tuning strategies. Overall, our investigation challenges the prevailing assumptions about effective representation learning in the online continual learning. | Random Representations Outperform Online Continually Learned Representations | [
"Ameya Prabhu",
"Shiven Sinha",
"Ponnurangam Kumaraguru",
"Philip Torr",
"Ozan Sener",
"Puneet K. Dokania"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=TYdzj1EvBP | @inproceedings{
chang2024how,
title={How Do Large Language Models Acquire Factual Knowledge During Pretraining?},
author={Hoyeon Chang and Jinho Park and Seonghyeon Ye and Sohee Yang and Youngkyung Seo and Du-Seong Chang and Minjoon Seo},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TYdzj1EvBP}
} | Despite the recent observation that large language models (LLMs) can store substantial factual knowledge, there is a limited understanding of the mechanisms of how they acquire factual knowledge through pretraining. This work addresses this gap by studying how LLMs acquire factual knowledge during pretraining. The findings reveal several important insights into the dynamics of factual knowledge acquisition during pretraining. First, counterintuitively, we observe that pretraining on more data shows no significant improvement in the model's capability to acquire and maintain factual knowledge. Next, LLMs undergo forgetting of memorization and generalization of factual knowledge, and LLMs trained with duplicated training data exhibit faster forgetting. Third, training LLMs with larger batch sizes can enhance the models' robustness to forgetting. Overall, our observations suggest that factual knowledge acquisition in LLM pretraining occurs by progressively increasing the probability of factual knowledge presented in the pretraining data at each step. However, this increase is diluted by subsequent forgetting. Based on this interpretation, we demonstrate that we can provide plausible explanations on recently observed behaviors of LLMs, such as the poor performance of LLMs on long-tail knowledge and the benefits of deduplicating the pretraining corpus. | How Do Large Language Models Acquire Factual Knowledge During Pretraining? | [
"Hoyeon Chang",
"Jinho Park",
"Seonghyeon Ye",
"Sohee Yang",
"Youngkyung Seo",
"Du-Seong Chang",
"Minjoon Seo"
] | NeurIPS.cc/2024/Conference | 2406.11813 | [
"https://github.com/kaistai/factual-knowledge-acquisition"
] | https://huggingface.co/papers/2406.11813 | 6 | 30 | 1 | 7 | [] | [
"kaist-ai/fictional-knowledge"
] | [] | [] | [
"kaist-ai/fictional-knowledge"
] | [] | 1 | poster |
null | https://openreview.net/forum?id=TY9VoSZZIA | @inproceedings{
datres2024a,
title={A two-scale Complexity Measure for Deep Learning Models},
author={Massimiliano Datres and Gian Paolo Leonardi and Alessio Figalli and David Sutter},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TY9VoSZZIA}
} | We introduce a novel capacity measure 2sED for statistical models based on the effective dimension. The new quantity provably bounds the generalization error under mild assumptions on the model. Furthermore, simulations on standard data sets and popular model architectures show that 2sED correlates well with the training error. For Markovian models, we show how to efficiently approximate 2sED from below through a layerwise iterative approach, which allows us to tackle deep learning models with a large number of parameters. Simulation results suggest that the approximation is good for different prominent models and data sets. | A two-scale Complexity Measure for Deep Learning Models | [
"Massimiliano Datres",
"Gian Paolo Leonardi",
"Alessio Figalli",
"David Sutter"
] | NeurIPS.cc/2024/Conference | 2401.09184 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TXsRGrzICz | @inproceedings{
lazaro-gredilla2024what,
title={What type of inference is planning?},
author={Miguel Lazaro-Gredilla and Li Yang Ku and Kevin Patrick Murphy and Dileep George},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TXsRGrzICz}
} | Multiple types of inference are available for probabilistic graphical models, e.g., marginal, maximum-a-posteriori, and even marginal maximum-a-posteriori. Which one do researchers mean when they talk about ``planning as inference''? There is no consistency in the literature, different types are used, and their ability to do planning is further entangled with specific approximations or additional constraints. In this work we use the variational framework to show that, just like all commonly used types of inference correspond to different weightings of the entropy terms in the variational problem, planning corresponds _exactly_ to a _different_ set of weights. This means that all the tricks of variational inference are readily applicable to planning. We develop an analogue of loopy belief propagation that allows us to perform approximate planning in factored-state Markov decisions processes without incurring intractability due to the exponentially large state space. The variational perspective shows that the previous types of inference for planning are only adequate in environments with low stochasticity, and allows us to characterize each type by its own merits, disentangling the type of inference from the additional approximations that its practical use requires. We validate these results empirically on synthetic MDPs and tasks posed in the International Planning Competition. | What type of inference is planning? | [
"Miguel Lazaro-Gredilla",
"Li Yang Ku",
"Kevin Patrick Murphy",
"Dileep George"
] | NeurIPS.cc/2024/Conference | 2406.17863 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=TWeVQ5meMW | @inproceedings{
miao2024subjectdriven,
title={Subject-driven Text-to-Image Generation via Preference-based Reinforcement Learning},
author={Yanting Miao and William Loh and Suraj Kothawade and Pascal Poupart and Abdullah Rashwan and Yeqing Li},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TWeVQ5meMW}
} | Text-to-image generative models have recently attracted considerable interest, enabling the synthesis of high-quality images from textual prompts. However, these models often lack the capability to generate specific subjects from given reference images or to synthesize novel renditions under varying conditions. Methods like DreamBooth and Subject-driven Text-to-Image (SuTI) have made significant progress in this area. Yet, both approaches primarily focus on enhancing similarity to reference images and require expensive setups, often overlooking the need for efficient training and avoiding overfitting to the reference images. In this work, we present the $\lambda$-Harmonic reward function, which provides a reliable reward signal and enables early stopping for faster training and effective regularization. By combining the Bradley-Terry preference model, the $\lambda$-Harmonic reward function also provides preference labels for subject-driven generation tasks. We propose Reward Preference Optimization (RPO), which offers a simpler setup (requiring only 3\% of the negative samples used by DreamBooth) and fewer gradient steps for fine-tuning. Unlike most existing methods, our approach does not require training a text encoder or optimizing text embeddings and achieves text-image alignment by fine-tuning only the U-Net component. Empirically, $\lambda$-Harmonic proves to be a reliable approach for model selection in subject-driven generation tasks. Based on preference labels and early stopping validation from the $\lambda$-Harmonic reward function, our algorithm achieves a state-of-the-art CLIP-I score of 0.833 and a CLIP-T score of 0.314 on DreamBench. | Subject-driven Text-to-Image Generation via Preference-based Reinforcement Learning | [
"Yanting Miao",
"William Loh",
"Suraj Kothawade",
"Pascal Poupart",
"Abdullah Rashwan",
"Yeqing Li"
] | NeurIPS.cc/2024/Conference | 2407.12164 | [
""
] | https://huggingface.co/papers/2407.12164 | 2 | 0 | 0 | 6 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=TVbCKAqoD8 | @inproceedings{
soen2024tradeoffs,
title={Trade-Offs of Diagonal Fisher Information Matrix Estimators},
author={Alexander Soen and Ke Sun},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TVbCKAqoD8}
} | The Fisher information matrix can be used to characterize the local geometry of
the parameter space of neural networks. It elucidates insightful theories and
useful tools to understand and optimize neural networks. Given its high
computational cost, practitioners often use random estimators and evaluate only
the diagonal entries. We examine two popular estimators whose accuracy and sample
complexity depend on their associated variances. We derive bounds of the
variances and instantiate them in neural networks for regression and
classification. We navigate trade-offs for both estimators based on analytical
and numerical studies. We find that the variance quantities depend on the
non-linearity w.r.t. different parameter groups and should not be neglected when
estimating the Fisher information. | Trade-Offs of Diagonal Fisher Information Matrix Estimators | [
"Alexander Soen",
"Ke Sun"
] | NeurIPS.cc/2024/Conference | 2402.05379 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TUwWBLjFk9 | @inproceedings{
xiang2024on,
title={On the Identifiability of Poisson Branching Structural Causal Model Using Probability Generating Function},
author={Yu Xiang and Jie Qiao and Zefeng Liang and Zihuai Zeng and Ruichu Cai and Zhifeng Hao},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TUwWBLjFk9}
} | Causal discovery from observational data, especially for count data, is essential across scientific and industrial contexts, such as biology, economics, and network operation maintenance. For this task, most approaches model count data using Bayesian networks or ordinal relations. However, they overlook the inherent branching structures that are frequently encountered, e.g., a browsing event might trigger an adding cart or purchasing event. This can be modeled by a binomial thinning operator (for branching) and an additive independent Poisson distribution (for noising), known as Poisson Branching Structure Causal Model (PB-SCM). There is a provably sound cumulant-based causal discovery method that allows the identification of the causal structure under a branching structure. However, we show that there still remains a gap in that there exist causal directions that are identifiable while the algorithm fails to identify them. In this work, we address this gap by exploring the identifiability of PB-SCM using the Probability Generating Function (PGF). By developing a compact and exact closed-form solution for the PGF of PB-SCM, we demonstrate that each component in this closed-form solution uniquely encodes a specific local structure, enabling the identification of the local structures by testing their corresponding component appearances in the PGF. Building on this, we propose a practical algorithm for learning causal skeletons and identifying causal directions of PB-SCM using PGF. The effectiveness of our method is demonstrated through experiments on both synthetic and real datasets. | On the Identifiability of Poisson Branching Structural Causal Model Using Probability Generating Function | [
"Yu Xiang",
"Jie Qiao",
"Zefeng Liang",
"Zihuai Zeng",
"Ruichu Cai",
"Zhifeng Hao"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
||
null | https://openreview.net/forum?id=TSaieShX3j | @inproceedings{
varre2024sgd,
title={{SGD} vs {GD}: Rank Deficiency in Linear Networks},
author={Aditya Varre and Margarita Sagitova and Nicolas Flammarion},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TSaieShX3j}
} | In this article, we study the behaviour of continuous-time gradient methods on a two-layer linear network with square loss. A dichotomy between SGD and GD is revealed: GD preserves the rank at initialization while (label noise) SGD diminishes the rank regardless of the initialization. We demonstrate this rank deficiency by studying the time evolution of the *determinant* of a matrix of parameters. To further understand this phenomenon, we derive the stochastic differential equation (SDE) governing the eigenvalues of the parameter matrix. This SDE unveils a *replusive force* between the eigenvalues: a key regularization mechanism which induces rank deficiency. Our results are well supported by experiments illustrating the phenomenon beyond linear networks and regression tasks. | SGD vs GD: Rank Deficiency in Linear Networks | [
"Aditya Varre",
"Margarita Sagitova",
"Nicolas Flammarion"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=TS09IypR3r | @inproceedings{
moreno2024metacurl,
title={Meta{CURL}: Non-stationary Concave Utility Reinforcement Learning},
author={Bianca Marin Moreno and Margaux Br{\'e}g{\`e}re and Pierre Gaillard and Nadia Oudjane},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TS09IypR3r}
} | We explore online learning in episodic loop-free Markov decision processes on non-stationary environments (changing losses and probability transitions). Our focus is on the Concave Utility Reinforcement Learning problem (CURL), an extension of classical RL for handling convex performance criteria in state-action distributions induced by agent policies. While various machine learning problems can be written as CURL, its non-linearity invalidates traditional Bellman equations. Despite recent solutions to classical CURL, none address non-stationary MDPs. This paper introduces MetaCURL, the first CURL algorithm for non-stationary MDPs. It employs a meta-algorithm running multiple black-box algorithms instances over different intervals, aggregating outputs via a sleeping expert framework. The key hurdle is partial information due to MDP uncertainty. Under partial information on the probability transitions (uncertainty and non-stationarity coming only from external noise, independent of agent state-action pairs), we achieve optimal dynamic regret without prior knowledge of MDP changes. Unlike approaches for RL, MetaCURL handles full adversarial losses, not just stochastic ones. We believe our approach for managing non-stationarity with experts can be of interest to the RL community. | MetaCURL: Non-stationary Concave Utility Reinforcement Learning | [
"Bianca Marin Moreno",
"Margaux Brégère",
"Pierre Gaillard",
"Nadia Oudjane"
] | NeurIPS.cc/2024/Conference | 2405.19807 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TPtXnpRvur | @inproceedings{
wu2024onestep,
title={One-Step Effective Diffusion Network for Real-World Image Super-Resolution},
author={Rongyuan Wu and Lingchen Sun and Zhiyuan Ma and Lei Zhang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TPtXnpRvur}
} | The pre-trained text-to-image diffusion models have been increasingly employed to tackle the real-world image super-resolution (Real-ISR) problem due to their powerful generative image priors. Most of the existing methods start from random noise to reconstruct the high-quality (HQ) image under the guidance of the given low-quality (LQ) image. While promising results have been achieved, such Real-ISR methods require multiple diffusion steps to reproduce the HQ image, increasing the computational cost. Meanwhile, the random noise introduces uncertainty in the output, which is unfriendly to image restoration tasks. To address these issues, we propose a one-step effective diffusion network, namely OSEDiff, for the Real-ISR problem.
We argue that the LQ image contains rich information to restore its HQ counterpart, and hence the given LQ image can be directly taken as the starting point for diffusion, eliminating the uncertainty introduced by random noise sampling. We finetune the pre-trained diffusion network with trainable layers to adapt it to complex image degradations. To ensure that the one-step diffusion model could yield HQ Real-ISR output, we apply variational score distillation in the latent space to conduct KL-divergence regularization. As a result, our OSEDiff model can efficiently and effectively generate HQ images in just one diffusion step.
Our experiments demonstrate that OSEDiff achieves comparable or even better Real-ISR results, in terms of both objective metrics and subjective evaluations, than previous diffusion model-based Real-ISR methods that require dozens or hundreds of steps. The source codes are released at https://github.com/cswry/OSEDiff. | One-Step Effective Diffusion Network for Real-World Image Super-Resolution | [
"Rongyuan Wu",
"Lingchen Sun",
"Zhiyuan Ma",
"Lei Zhang"
] | NeurIPS.cc/2024/Conference | [
"https://github.com/cswry/osediff"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=TNQ0hxh3O1 | @inproceedings{
huang2024openvocabulary,
title={Open-Vocabulary Object Detection via Language Hierarchy},
author={Jiaxing Huang and Jingyi Zhang and Kai Jiang and Shijian Lu},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TNQ0hxh3O1}
} | Recent studies on generalizable object detection have attracted increasing attention with additional weak supervision from large-scale datasets with image-level labels.
However, weakly-supervised detection learning often suffers from image-to-box label mismatch, i.e., image-level
labels do not convey precise object information.
We design Language Hierarchical Self-training (LHST) that introduces language hierarchy into weakly-supervised detector training for learning more generalizable detectors.
LHST expands the image-level labels with language hierarchy and enables co-regularization between the expanded labels and self-training. Specifically, the expanded labels regularize self-training by providing richer supervision and mitigating the image-to-box label mismatch, while self-training allows assessing and selecting the expanded labels according to the predicted reliability.
In addition, we design language hierarchical prompt generation that introduces language hierarchy into prompt generation which helps bridge the vocabulary gaps between training and testing.
Extensive experiments show that the proposed techniques achieve superior generalization performance consistently across 14 widely studied object detection datasets. | Open-Vocabulary Object Detection via Language Hierarchy | [
"Jiaxing Huang",
"Jingyi Zhang",
"Kai Jiang",
"Shijian Lu"
] | NeurIPS.cc/2024/Conference | 2410.20371 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TNEmAgwoXR | @inproceedings{
tian2024confident,
title={Confident Natural Policy Gradient for Local Planning in \$q\_{\textbackslash}pi\$-realizable Constrained {MDP}s},
author={Tian Tian and Lin Yang and Csaba Szepesvari},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TNEmAgwoXR}
} | The constrained Markov decision process (CMDP) framework emerges as an important reinforcement learning approach for imposing safety or other critical objectives while maximizing cumulative reward. However, the current understanding of how to learn efficiently in a CMDP environment with a potentially infinite number of states remains under investigation, particularly when function approximation is applied to the value functions. In this paper, we address the learning problem given linear function approximation with $q_{\pi}$-realizability, where the value functions of all policies are linearly representable with a known feature map, a setting known to be more general and challenging than other linear settings. Utilizing a local-access model, we propose a novel primal-dual algorithm that, after $\tilde{O}(\text{poly}(d) \epsilon^{-3})$ iterations, outputs with high probability a policy that strictly satisfies the constraints while nearly optimizing the value with respect to a reward function. Here, $d$ is the feature dimension and $\epsilon > 0$ is a given error. The algorithm relies on a carefully crafted off-policy evaluation procedure to evaluate the policy using historical data, which informs policy updates through policy gradients and conserves samples. To our knowledge, this is the first result achieving polynomial sample complexity for CMDP in the $q_{\pi}$-realizable setting. | Confident Natural Policy Gradient for Local Planning in q_π-realizable Constrained MDPs | [
"Tian Tian",
"Lin Yang",
"Csaba Szepesvari"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=TMlGQw7EbC | @inproceedings{
deng2024markov,
title={Markov Equivalence and Consistency in Differentiable Structure Learning},
author={Chang Deng and Kevin Bello and Pradeep Kumar Ravikumar and Bryon Aragam},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TMlGQw7EbC}
} | Existing approaches to differentiable structure learning of directed acyclic graphs (DAGs) rely on strong identifiability assumptions in order to guarantee that global minimizers of the acyclicity-constrained optimization problem identifies the true DAG. Moreover, it has been observed empirically that the optimizer may exploit undesirable artifacts in the loss function. We explain and remedy these issues by studying the behavior of differentiable acyclicity-constrained programs under general likelihoods with multiple global minimizers. By carefully regularizing the likelihood, it is possible to identify the sparsest model in the Markov equivalence class, even in the absence of an identifiable parametrization. We first study the Gaussian case in detail, showing how proper regularization of the likelihood defines a score that identifies the sparsest model. Assuming faithfulness, it also recovers the Markov equivalence class. These results are then generalized to general models and likelihoods, where the same claims hold. These theoretical results are validated empirically, showing how this can be done using standard gradient-based optimizers, thus paving the way for differentiable structure learning under general models and losses. | Markov Equivalence and Consistency in Differentiable Structure Learning | [
"Chang Deng",
"Kevin Bello",
"Pradeep Kumar Ravikumar",
"Bryon Aragam"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=TLUGoShY30 | @inproceedings{
tengjie2024multitimes,
title={Multi-times Monte Carlo Rendering for Inter-reflection Reconstruction},
author={Zhu Tengjie and Zhuo Chen and Jingnan Gao and Yichao Yan and Xiaokang Yang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TLUGoShY30}
} | Inverse rendering methods have achieved remarkable performance in reconstructing high-fidelity 3D objects with disentangled geometries, materials, and environmental light. However, they still face huge challenges in reflective surface reconstruction. Although recent methods model the light trace to learn specularity, the ignorance of indirect illumination makes it hard to handle inter-reflections among multiple smooth objects. In this work, we propose Ref-MC2 that introduces the multi-time Monte Carlo sampling which comprehensively computes the environmental illumination and meanwhile considers the reflective light from object surfaces. To address the computation challenge as the times of Monte Carlo sampling grow, we propose a specularity-adaptive sampling strategy, significantly reducing the computational complexity. Besides the computational resource, higher geometry accuracy is also required because geometric errors accumulate multiple times. Therefore, we further introduce a reflection-aware surface model to initialize the geometry and refine it during inverse rendering. We construct a challenging dataset containing scenes with multiple objects and inter-reflections. Experiments show that our method outperforms other inverse rendering methods on various object groups. We also show downstream applications, e.g., relighting and material editing, to illustrate the disentanglement ability of our method. | Multi-times Monte Carlo Rendering for Inter-reflection Reconstruction | [
"Zhu Tengjie",
"Zhuo Chen",
"Jingnan Gao",
"Yichao Yan",
"Xiaokang Yang"
] | NeurIPS.cc/2024/Conference | 2407.05771 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TJsknGasMy | @inproceedings{
birrell2024differentially,
title={Differentially Private Stochastic Gradient Descent with Fixed-Size Minibatches: Tighter {RDP} Guarantees with or without Replacement},
author={Jeremiah Birrell and Mohammadreza Ebrahimi and Rouzbeh Behnia and Jason Pacheco},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TJsknGasMy}
} | Differentially private stochastic gradient descent (DP-SGD) has been instrumental in privately training deep learning models by providing a framework to control and track the privacy loss incurred during training. At the core of this computation lies a subsampling method that uses a privacy amplification lemma to enhance the privacy guarantees provided by the additive noise. Fixed size subsampling is appealing for its constant memory usage, unlike the variable sized minibatches in Poisson subsampling. It is also of interest in addressing class imbalance and federated learning. Current computable guarantees for fixed-size subsampling are not tight and do not consider both add/remove and replace-one adjacency relationships. We present a new and holistic Rényi differential privacy (RDP) accountant for DP-SGD with fixed-size subsampling without replacement (FSwoR) and with replacement (FSwR). For FSwoR we consider both add/remove and replace-one adjacency, where we improve on the best current computable bound by a factor of $4$. We also show for the first time that the widely-used Poisson subsampling and FSwoR with replace-one adjacency have the same privacy to leading order in the sampling probability. Our work suggests that FSwoR is often preferable to Poisson subsampling due to constant memory usage. Our FSwR accountant includes explicit non-asymptotic upper and lower bounds and, to the authors' knowledge, is the first such RDP analysis of fixed-size subsampling with replacement for DP-SGD. We analytically and empirically compare fixed size and Poisson subsampling, and show that DP-SGD gradients in a fixed-size subsampling regime exhibit lower variance in practice in addition to memory usage benefits. | Differentially Private Stochastic Gradient Descent with Fixed-Size Minibatches: Tighter RDP Guarantees with or without Replacement | [
"Jeremiah Birrell",
"Mohammadreza Ebrahimi",
"Rouzbeh Behnia",
"Jason Pacheco"
] | NeurIPS.cc/2024/Conference | 2408.10456 | [
"https://github.com/star-ailab/FSRDP"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TJiw1oLAcD | @inproceedings{
huang2024improving,
title={Improving Viewpoint-Independent Object-Centric Representations through Active Viewpoint Selection},
author={Yinxuan Huang and Chengmin Gao and Bin Li and Xiangyang Xue},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TJiw1oLAcD}
} | Given the complexities inherent in visual scenes, such as object occlusion, a comprehensive understanding often requires observation from multiple viewpoints. Existing multi-viewpoint object-centric learning methods typically employ random or sequential viewpoint selection strategies. While applicable across various scenes, these strategies may not always be ideal, as certain scenes could benefit more from specific viewpoints. To address this limitation, we propose a novel active viewpoint selection strategy. This strategy predicts images from unknown viewpoints based on information from observation images for each scene. It then compares the object-centric representations extracted from both viewpoints and selects the unknown viewpoint with the largest disparity, indicating the greatest gain in information, as the next observation viewpoint. Through experiments on various datasets, we demonstrate the effectiveness of our active viewpoint selection strategy, significantly enhancing segmentation and reconstruction performance compared to random viewpoint selection. Moreover, our method can accurately predict images from unknown viewpoints. | Improving Viewpoint-Independent Object-Centric Representations through Active Viewpoint Selection | [
"Yinxuan Huang",
"Chengmin Gao",
"Bin Li",
"Xiangyang Xue"
] | NeurIPS.cc/2024/Conference | 2411.00402 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TIhiFqGOYC | @inproceedings{
xiong2024meaningful,
title={Meaningful Learning: Enhancing Abstract Reasoning in Large Language Models via Generic Fact Guidance},
author={Kai Xiong and Xiao Ding and Ting Liu and Bing Qin and Dongliang Xu and Qing Yang and Hongtao Liu and Yixin Cao},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TIhiFqGOYC}
} | Large language models (LLMs) have developed impressive performance and strong explainability across various reasoning scenarios, marking a significant stride towards mimicking human-like intelligence. Despite this, when tasked with several simple questions supported by a generic fact, LLMs often struggle to abstract and apply the generic fact to provide consistent and precise answers, revealing a deficiency in abstract reasoning abilities. This has sparked a vigorous debate about whether LLMs are genuinely reasoning or merely memorizing. In light of this, we design a preliminary study to quantify and delve into the abstract reasoning abilities of existing LLMs. Our findings reveal a substantial discrepancy between their general reasoning and abstract reasoning performances. To relieve this problem, we tailor an abstract reasoning dataset (AbsR) together with a meaningful learning paradigm to teach LLMs how to leverage generic facts for reasoning purposes. The results show that our approach not only boosts the general reasoning performance of LLMs but also makes considerable strides towards their capacity for abstract reasoning, moving beyond simple memorization or imitation to a more nuanced understanding and application of generic facts. The code is available at https://github.com/Waste-Wood/MeanLearn. | Meaningful Learning: Enhancing Abstract Reasoning in Large Language Models via Generic Fact Guidance | [
"Kai Xiong",
"Xiao Ding",
"Ting Liu",
"Bing Qin",
"Dongliang Xu",
"Qing Yang",
"Hongtao Liu",
"Yixin Cao"
] | NeurIPS.cc/2024/Conference | 2403.09085 | [
"https://github.com/waste-wood/meanlearn"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TGmwp9jJXl | @inproceedings{
devergne2024from,
title={From Biased to Unbiased Dynamics: An Infinitesimal Generator Approach},
author={Timoth{\'e}e Devergne and Vladimir R Kostic and Michele Parrinello and Massimiliano Pontil},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TGmwp9jJXl}
} | We investigate learning the eigenfunctions of evolution operators for time-reversal invariant stochastic processes, a prime example being the Langevin equation used in molecular dynamics. Many physical or chemical processes described by this equation involve transitions between metastable states separated by high potential barriers that can hardly be crossed during a simulation. To overcome this bottleneck, data are collected via biased simulations that explore the state space more rapidly. We propose a framework for learning from biased simulations rooted in the infinitesimal generator of the process {and the associated resolvent operator}. We contrast our approach to more common ones based on the transfer operator, showing that it can provably learn the spectral properties of the unbiased system from biased data. In experiments, we highlight the advantages of our method over transfer operator approaches and recent developments based on generator learning, demonstrating its effectiveness in estimating eigenfunctions and eigenvalues. Importantly, we show that even with datasets containing only a few relevant transitions due to sub-optimal biasing, our approach
recovers relevant information about the transition mechanism. | From Biased to Unbiased Dynamics: An Infinitesimal Generator Approach | [
"Timothée Devergne",
"Vladimir R Kostic",
"Michele Parrinello",
"Massimiliano Pontil"
] | NeurIPS.cc/2024/Conference | 2406.09028 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TGC7HNf6nK | @inproceedings{
yang2024lever,
title={Lever {LM}: Configuring In-Context Sequence to Lever Large Vision Language Models},
author={Xu Yang and Yingzhe Peng and Haoxuan Ma and Shuo Xu and Chi Zhang and Yucheng Han and Hanwang Zhang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TGC7HNf6nK}
} | As Archimedes famously said, ``Give me a lever long enough and a fulcrum on which to place it, and I shall move the world'', in this study, we propose to use a tiny Language Model (LM), \eg, a Transformer with 67M parameters, to lever much larger Vision-Language Models (LVLMs) with 9B parameters. Specifically, we use this tiny \textbf{Lever-LM} to configure effective in-context demonstration (ICD) sequences to improve the In-Context Learinng (ICL) performance of LVLMs. Previous studies show that diverse ICD configurations like the selection and ordering of the demonstrations heavily affect the ICL performance, highlighting the significance of configuring effective ICD sequences. Motivated by this and by re-considering the the process of configuring ICD sequence, we find this is a mirror process of human sentence composition and further assume that effective ICD configurations may contain internal statistical patterns that can be captured by Lever-LM. Then a dataset with effective ICD sequences is constructed to train Lever-LM. After training, given novel queries, new ICD sequences are configured by the trained Lever-LM to solve vision-language tasks through ICL. Experiments show that these ICD sequences can improve the ICL performance of two LVLMs compared with some strong baselines in Visual Question Answering and Image Captioning, validating that Lever-LM can really capture the statistical patterns for levering LVLMs. The code is available at \url{https://anonymous.4open.science/r/Lever-LM-604A/}. | Lever LM: Configuring In-Context Sequence to Lever Large Vision Language Models | [
"Xu Yang",
"Yingzhe Peng",
"Haoxuan Ma",
"Shuo Xu",
"Chi Zhang",
"Yucheng Han",
"Hanwang Zhang"
] | NeurIPS.cc/2024/Conference | 2312.10104 | [
"https://github.com/forjadeforest/icd-lm"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=TFZlFRl9Ks | @inproceedings{
gao2024catd,
title={{CAT}3D: Create Anything in 3D with Multi-View Diffusion Models},
author={Ruiqi Gao and Aleksander Holynski and Philipp Henzler and Arthur Brussee and Ricardo Martin Brualla and Pratul P. Srinivasan and Jonathan T. Barron and Ben Poole},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TFZlFRl9Ks}
} | Advances in 3D reconstruction have enabled high-quality 3D capture, but require a user to collect hundreds to thousands of images to create a 3D scene. We present CAT3D, a method for creating anything in 3D by simulating this real-world capture process with a multi-view diffusion model. Given any number of input images and a set of target novel viewpoints, our model generates highly consistent novel views of a scene. These generated views can be used as input to robust 3D reconstruction techniques to produce 3D representations that can be rendered from any viewpoint in real-time. CAT3D can create entire 3D scenes in as little as one minute, and outperforms existing methods for single image and few-view 3D scene creation. | CAT3D: Create Anything in 3D with Multi-View Diffusion Models | [
"Ruiqi Gao",
"Aleksander Holynski",
"Philipp Henzler",
"Arthur Brussee",
"Ricardo Martin Brualla",
"Pratul P. Srinivasan",
"Jonathan T. Barron",
"Ben Poole"
] | NeurIPS.cc/2024/Conference | 2405.10314 | [
""
] | https://huggingface.co/papers/2405.10314 | 4 | 44 | 2 | 8 | [] | [] | [] | [] | [] | [] | 1 | oral |
null | https://openreview.net/forum?id=TFAG9UznPv | @inproceedings{
altstidl2024on,
title={On the Scalability of Certified Adversarial Robustness with Generated Data},
author={Thomas Altstidl and David Dobre and Arthur Kosmala and Bjoern Eskofier and Gauthier Gidel and Leo Schwinn},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TFAG9UznPv}
} | Certified defenses against adversarial attacks offer formal guarantees on the robustness of a model, making them more reliable than empirical methods such as adversarial training, whose effectiveness is often later reduced by unseen attacks. Still, the limited certified robustness that is currently achievable has been a bottleneck for their practical adoption. Gowal et al. and Wang et al. have shown that generating additional training data using state-of-the-art diffusion models can considerably improve the robustness of adversarial training. In this work, we demonstrate that a similar approach can substantially improve deterministic certified defenses but also reveal notable differences in the scaling behavior between certified and empirical methods. In addition, we provide a list of recommendations to scale the robustness of certified training approaches. Our approach achieves state-of-the-art deterministic robustness certificates on CIFAR-10 for the $\ell_2$ ($\epsilon = 36/255$) and $\ell_{\infty}$ ($\epsilon = 8/255$) threat models, outperforming the previous results by $+3.95$ and $+1.39$ percentage points, respectively. Furthermore, we report similar improvements for CIFAR-100. | On the Scalability of Certified Adversarial Robustness with Generated Data | [
"Thomas Altstidl",
"David Dobre",
"Arthur Kosmala",
"Bjoern Eskofier",
"Gauthier Gidel",
"Leo Schwinn"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=TBVLQjdFcA | @inproceedings{
wei2024generated,
title={Generated and Pseudo Content guided Prototype Refinement for Few-shot Point Cloud Segmentation},
author={Lili Wei and Congyan Lang and Ziyi Chen and Tao Wang and Yidong Li and Jun Liu},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TBVLQjdFcA}
} | Few-shot 3D point cloud semantic segmentation aims to segment query point clouds with only a few annotated support point clouds. Existing prototype-based methods learn prototypes from the 3D support set to guide the segmentation of query point clouds. However, they encounter the challenge of low prototype quality due to constrained semantic information in the 3D support set and class information bias between support and query sets. To address these issues, in this paper, we propose a novel framework called Generated and Pseudo Content guided Prototype Refinement (GPCPR), which explicitly leverages LLM-generated content and reliable query context to enhance prototype quality. GPCPR achieves prototype refinement through two core components: LLM-driven Generated Content-guided Prototype Refinement (GCPR) and Pseudo Query Context-guided Prototype Refinement (PCPR). Specifically, GCPR integrates diverse and differentiated class descriptions generated by large language models to enrich prototypes with comprehensive semantic knowledge. PCPR further aggregates reliable class-specific pseudo-query context to mitigate class information bias and generate more suitable query-specific prototypes. Furthermore, we introduce a dual-distillation regularization term, enabling knowledge transfer between early-stage entities (prototypes or pseudo predictions) and their deeper counterparts to enhance refinement. Extensive experiments demonstrate the superiority of our method, surpassing the state-of-the-art methods by up to 12.10% and 13.75% mIoU on S3DIS and ScanNet, respectively. | Generated and Pseudo Content guided Prototype Refinement for Few-shot Point Cloud Segmentation | [
"Lili Wei",
"Congyan Lang",
"Ziyi Chen",
"Tao Wang",
"Yidong Li",
"Jun Liu"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
||
null | https://openreview.net/forum?id=TALJtWX7w4 | @inproceedings{
popordanoska2024lascal,
title={La{SC}al: Label-Shift Calibration without target labels},
author={Teodora Popordanoska and Gorjan Radevski and Tinne Tuytelaars and Matthew B. Blaschko},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TALJtWX7w4}
} | When machine learning systems face dataset shift, model calibration plays a pivotal role in ensuring their reliability.
Calibration error (CE) provides insights into the alignment between the predicted confidence scores and the classifier accuracy.
While prior works have delved into the implications of dataset shift on calibration, existing CE estimators either (i) assume access to labeled data from the target domain, often unavailable in practice, or (ii) are derived under a covariate shift assumption.
In this work we propose a novel, label-free, consistent CE estimator under label shift. Label shift is characterized by changes in the marginal label distribution p(Y), with a constant conditional p(X|Y) distribution between the source and target. We introduce a novel calibration method, called LaSCal, which uses the estimator in conjunction with a post-hoc calibration strategy, to perform unsupervised calibration on the target distribution. Our thorough empirical analysis demonstrates the effectiveness and reliability of the proposed approach across different modalities, model architectures and label shift intensities. | LaSCal: Label-Shift Calibration without target labels | [
"Teodora Popordanoska",
"Gorjan Radevski",
"Tinne Tuytelaars",
"Matthew B. Blaschko"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=TADTT9ughN | @inproceedings{
melo2024deep,
title={Deep Bayesian Active Learning for Preference Modeling in Large Language Models},
author={Luckeciano Carvalho Melo and Panagiotis Tigas and Alessandro Abate and Yarin Gal},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TADTT9ughN}
} | Leveraging human preferences for steering the behavior of Large Language Models (LLMs) has demonstrated notable success in recent years. Nonetheless, data selection and labeling are still a bottleneck for these systems, particularly at large scale. Hence, selecting the most informative points for acquiring human feedback may considerably reduce the cost of preference labeling and unleash the further development of LLMs. Bayesian Active Learning provides a principled framework for addressing this challenge and has demonstrated remarkable success in diverse settings. However, previous attempts to employ it for Preference Modeling did not meet such expectations. In this work, we identify that naive epistemic uncertainty estimation leads to the acquisition of redundant samples. We address this by proposing the Bayesian Active Learner for Preference Modeling (BAL-PM), a novel stochastic acquisition policy that not only targets points of high epistemic uncertainty according to the preference model but also seeks to maximize the entropy of the acquired prompt distribution in the feature space spanned by the employed LLM. Notably, our experiments demonstrate that BAL-PM requires 33\% to 68\% fewer preference labels in two popular human preference datasets and exceeds previous stochastic Bayesian acquisition policies. | Deep Bayesian Active Learning for Preference Modeling in Large Language Models | [
"Luckeciano Carvalho Melo",
"Panagiotis Tigas",
"Alessandro Abate",
"Yarin Gal"
] | NeurIPS.cc/2024/Conference | 2406.10023 | [
"https://github.com/luckeciano/bal-pm"
] | https://huggingface.co/papers/2406.10023 | 1 | 2 | 1 | 4 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=TA5zPfH8iI | @inproceedings{
arya2024bcosification,
title={B-cosification: Transforming Deep Neural Networks to be Inherently Interpretable},
author={Shreyash Arya and Sukrut Rao and Moritz B{\"o}hle and Bernt Schiele},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=TA5zPfH8iI}
} | B-cos Networks have been shown to be effective for obtaining highly human interpretable explanations of model decisions by architecturally enforcing stronger alignment between inputs and weight. B-cos variants of convolutional networks (CNNs) and vision transformers (ViTs), which primarily replace linear layers with B-cos transformations, perform competitively to their respective standard variants while also yielding explanations that are faithful by design. However, it has so far been necessary to train these models from scratch, which is increasingly infeasible in the era of large, pre-trained foundation models. In this work, inspired by the architectural similarities in standard DNNs and B-cos networks, we propose ‘B-cosification’, a novel approach to transform existing pre-trained models to become inherently interpretable. We perform a thorough study of design choices to perform this conversion, both for convolutional neural networks and vision transformers. We find that B-cosification can yield models that are on par with B-cos models trained from scratch in terms of interpretability, while often outperforming them in terms of classification performance at a fraction of the training cost. Subsequently, we apply B-cosification to a pretrained CLIP model, and show that, even with limited data and compute cost, we obtain a B-cosified version that is highly interpretable and competitive on zero shot performance across a variety of datasets. We release our
code and pre-trained model weights at https://github.com/shrebox/B-cosification. | B-cosification: Transforming Deep Neural Networks to be Inherently Interpretable | [
"Shreyash Arya",
"Sukrut Rao",
"Moritz Böhle",
"Bernt Schiele"
] | NeurIPS.cc/2024/Conference | 2411.00715 | [
"https://github.com/shrebox/b-cosification"
] | https://huggingface.co/papers/2411.00715 | 2 | 0 | 0 | 4 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=T9PfJViMiJ | @inproceedings{
xu2024hhdgp,
title={{HHD}-{GP}: Incorporating Helmholtz-Hodge Decomposition into Gaussian Processes for Learning Dynamical Systems},
author={Hao Xu and Jia Pan},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T9PfJViMiJ}
} | Machine learning models provide alternatives for efficiently recognizing complex patterns from data, but the main concern in applying them to modeling physical systems stems from their physics-agnostic design, leading to learning methods that lack interpretability, robustness, and data efficiency. This paper mitigates this concern by incorporating the Helmholtz-Hodge decomposition into a Gaussian process model, leading to a versatile framework that simultaneously learns the curl-free and divergence-free components of a dynamical system. Learning a predictive model in this form facilitates the exploitation of symmetry priors. In addition to improving predictive power, these priors make the model indentifiable, thus the identified features can be linked to comprehensible scientific properties of the system. We show that compared to baseline models, our model achieves better predictive performance on several benchmark dynamical systems while allowing physically meaningful decomposition of the systems from noisy and sparse data. | HHD-GP: Incorporating Helmholtz-Hodge Decomposition into Gaussian Processes for Learning Dynamical Systems | [
"Hao Xu",
"Jia Pan"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=T9GbbWbNQG | @inproceedings{
gwak2024layeradaptive,
title={Layer-Adaptive State Pruning for Deep State Space Models},
author={Minseon Gwak and Seongrok Moon and Joohwan Ko and PooGyeon Park},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T9GbbWbNQG}
} | Due to the lack of state dimension optimization methods, deep state space models (SSMs) have sacrificed model capacity, training search space, or stability to alleviate computational costs caused by high state dimensions. In this work, we provide a structured pruning method for SSMs, Layer-Adaptive STate pruning (LAST), which reduces the state dimension of each layer in minimizing model-level energy loss by extending modal truncation for a single system. LAST scores are evaluated using $\mathcal{H}_{\infty}$ norms of subsystems for each state and layer-wise energy normalization. The scores serve as global pruning criteria, enabling cross-layer comparison of states and layer-adaptive pruning. Across various sequence benchmarks, LAST optimizes previous SSMs, revealing the redundancy and compressibility of their state spaces. Notably, we demonstrate that, on average, pruning 33\% of states still maintains performance with 0.52\% accuracy loss in multi-input multi-output SSMs without retraining. Code is available at https://github.com/msgwak/LAST. | Layer-Adaptive State Pruning for Deep State Space Models | [
"Minseon Gwak",
"Seongrok Moon",
"Joohwan Ko",
"PooGyeon Park"
] | NeurIPS.cc/2024/Conference | 2411.02824 | [
"https://github.com/msgwak/last"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=T826pwZLci | @inproceedings{
gao2024federated,
title={Federated Online Prediction from Experts with Differential Privacy: Separations and Regret Speed-ups},
author={Fengyu Gao and Ruiquan Huang and Jing Yang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T826pwZLci}
} | We study the problems of differentially private federated online prediction from experts against both *stochastic adversaries* and *oblivious adversaries*. We aim to minimize the average regret on $m$ clients working in parallel over time horizon $T$ with explicit differential privacy (DP) guarantees. With stochastic adversaries, we propose a **Fed-DP-OPE-Stoch** algorithm that achieves $\sqrt{m}$-fold speed-up of the per-client regret compared to the single-player counterparts under both pure DP and approximate DP constraints, while maintaining logarithmic communication costs. With oblivious adversaries, we establish non-trivial lower bounds indicating that *collaboration among clients does not lead to regret speed-up with general oblivious adversaries*. We then consider a special case of the oblivious adversaries setting, where there exists a low-loss expert. We design a new algorithm **Fed-SVT** and show that it achieves an $m$-fold regret speed-up under both pure DP and approximate DP constraints over the single-player counterparts. Our lower bound indicates that Fed-SVT is nearly optimal up to logarithmic factors. Experiments demonstrate the effectiveness of our proposed algorithms. To the best of our knowledge, this is the first work examining the differentially private online prediction from experts in the federated setting. | Federated Online Prediction from Experts with Differential Privacy: Separations and Regret Speed-ups | [
"Fengyu Gao",
"Ruiquan Huang",
"Jing Yang"
] | NeurIPS.cc/2024/Conference | 2409.19092 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=T7dS1Ghwwu | @inproceedings{
shi2024conformal,
title={Conformal Prediction for Class-wise Coverage via Augmented Label Rank Calibration},
author={Yuanjie Shi and SUBHANKAR GHOSH and Taha Belkhouja and Jana Doppa and Yan Yan},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T7dS1Ghwwu}
} | Conformal prediction (CP) is an emerging uncertainty quantification framework that allows us to construct a prediction set to cover the true label with a pre-specified marginal or conditional probability.
Although the valid coverage guarantee has been extensively studied for classification problems, CP often produces large prediction sets which may not be practically useful.
This issue is exacerbated for the setting of class-conditional coverage on imbalanced classification tasks with many and/or imbalanced classes.
This paper proposes the Rank Calibrated Class-conditional CP (RC3P) algorithm to reduce the prediction set sizes to achieve class-conditional coverage, where the valid coverage holds for each class.
In contrast to the standard class-conditional CP (CCP) method that uniformly thresholds the class-wise conformity score for each class, the augmented label rank calibration step allows RC3P to selectively iterate this class-wise thresholding subroutine only for a subset of classes whose class-wise top-$k$ error is small.
We prove that agnostic to the classifier and data distribution, RC3P achieves class-wise coverage. We also show that RC3P reduces the size of prediction sets compared to the CCP method.
Comprehensive experiments on multiple real-world datasets demonstrate that RC3P achieves class-wise coverage and $26.25\\%$ $\downarrow$ reduction in prediction set sizes on average. | Conformal Prediction for Class-wise Coverage via Augmented Label Rank Calibration | [
"Yuanjie Shi",
"SUBHANKAR GHOSH",
"Taha Belkhouja",
"Jana Doppa",
"Yan Yan"
] | NeurIPS.cc/2024/Conference | 2406.06818 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=T6LOGZBC2m | @inproceedings{
nie2024opera,
title={{OPERA}: Automatic Offline Policy Evaluation with Re-weighted Aggregates of Multiple Estimators},
author={Allen Nie and Yash Chandak and Christina J. Yuan and Anirudhan Badrinath and Yannis Flet-Berliac and Emma Brunskill},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T6LOGZBC2m}
} | Offline policy evaluation (OPE) allows us to evaluate and estimate a new sequential decision-making policy's performance by leveraging historical interaction data collected from other policies. Evaluating a new policy online without a confident estimate of its performance can lead to costly, unsafe, or hazardous outcomes, especially in education and healthcare. Several OPE estimators have been proposed in the last decade, many of which have hyperparameters and require training. Unfortunately, choosing the best OPE algorithm for each task and domain is still unclear. In this paper, we propose a new algorithm that adaptively blends a set of OPE estimators given a dataset without relying on an explicit selection using a statistical procedure. We prove that our estimator is consistent and satisfies several desirable properties for policy evaluation. Additionally, we demonstrate that when compared to alternative approaches, our estimator can be used to select higher-performing policies in healthcare and robotics. Our work contributes to improving ease of use for a general-purpose, estimator-agnostic, off-policy evaluation framework for offline RL. | OPERA: Automatic Offline Policy Evaluation with Re-weighted Aggregates of Multiple Estimators | [
"Allen Nie",
"Yash Chandak",
"Christina J. Yuan",
"Anirudhan Badrinath",
"Yannis Flet-Berliac",
"Emma Brunskill"
] | NeurIPS.cc/2024/Conference | 2405.17708 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=T5UfIfmDbq | @inproceedings{
wang2024monte,
title={Monte Carlo Tree Search based Space Transfer for Black Box Optimization},
author={Shukuan Wang and Ke Xue and Lei Song and Xiaobin Huang and Chao Qian},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T5UfIfmDbq}
} | Bayesian optimization (BO) is a popular method for computationally expensive black-box optimization. However, traditional BO methods need to solve new problems from scratch, leading to slow convergence. Recent studies try to extend BO to a transfer learning setup to speed up the optimization, where search space transfer is one of the most promising approaches and has shown impressive performance on many tasks. However, existing search space transfer methods either lack an adaptive mechanism or are not flexible enough, making it difficult to efficiently identify promising search space during the optimization process. In this paper, we propose a search space transfer learning method based on Monte Carlo tree search (MCTS), called MCTS-transfer, to iteratively divide, select, and optimize in a learned subspace. MCTS-transfer can not only provide a well-performing search space for warm-start but also adaptively identify and leverage the information of similar source tasks to reconstruct the search space during the optimization process. Experiments on synthetic functions, real-world problems, Design-Bench and hyper-parameter optimization show that MCTS-transfer can demonstrate superior performance compared to other search space transfer methods under different settings. Our code is available at \url{https://github.com/lamda-bbo/mcts-transfer}. | Monte Carlo Tree Search based Space Transfer for Black Box Optimization | [
"Shukuan Wang",
"Ke Xue",
"Lei Song",
"Xiaobin Huang",
"Chao Qian"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
||
null | https://openreview.net/forum?id=T5Cerv7PT2 | @inproceedings{
hugessen2024simplifying,
title={Simplifying Constraint Inference with Inverse Reinforcement Learning},
author={Adriana Hugessen and Harley Wiltzer and Glen Berseth},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T5Cerv7PT2}
} | Learning safe policies has presented a longstanding challenge for the reinforcement learning (RL) community. Various formulations of safe RL have been proposed; However, fundamentally, tabula rasa RL must learn safety constraints through experience, which is problematic for real-world applications. Imitation learning is often preferred in real-world settings because the experts' safety preferences are embedded in the data the agent imitates. However, imitation learning is limited in its extensibility to new tasks, which can only be learned by providing the agent with expert trajectories. For safety-critical applications with sub-optimal or inexact expert data, it would be preferable to learn only the safety aspects of the policy through imitation, while still allowing for task learning with RL. The field of inverse constrained RL, which seeks to infer constraints from expert data, is a promising step in this direction. However, prior work in this area has relied on complex tri-level optimizations in order to infer safe behavior (constraints). This challenging optimization landscape leads to sub-optimal performance on several benchmark tasks. In this work, we present a simplified version of constraint inference that performs as well or better than prior work across a collection of continuous-control benchmarks. Moreover, besides improving performance, this simplified framework is easier to implement, tune, and more readily lends itself to various extensions, such as offline constraint inference. | Simplifying Constraint Inference with Inverse Reinforcement Learning | [
"Adriana Hugessen",
"Harley Wiltzer",
"Glen Berseth"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=T56j6aV8Oc | @inproceedings{
kunstner2024heavytailed,
title={Heavy-Tailed Class Imbalance and Why Adam Outperforms Gradient Descent on Language Models},
author={Frederik Kunstner and Robin Yadav and Alan Milligan and Mark Schmidt and Alberto Bietti},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T56j6aV8Oc}
} | Adam has been shown to outperform gradient descent on large language models by a larger margin than on other tasks, but it is unclear why. We show that a key factor in this performance gap is the heavy-tailed class imbalance found in language tasks. When trained with gradient descent, the loss of infrequent words decreases more slowly than the loss of frequent ones. This leads to a slow decrease on the average loss as most samples come from infrequent words. On the other hand, Adam and sign-based methods are less sensitive to this problem. To establish that this behavior is caused by class imbalance, we show empirically that it can be reproduced across architectures and data types, on language transformers, vision CNNs, and linear models. On a linear model with cross-entropy loss, we show that class imbalance leads to imbalanced, correlated gradients and Hessians that have been hypothesized to benefit Adam. We also prove that, in continuous time, gradient descent converges slowly on low-frequency classes while sign descent does not. | Heavy-Tailed Class Imbalance and Why Adam Outperforms Gradient Descent on Language Models | [
"Frederik Kunstner",
"Robin Yadav",
"Alan Milligan",
"Mark Schmidt",
"Alberto Bietti"
] | NeurIPS.cc/2024/Conference | 2402.19449 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=T1lFrYwtf7 | @inproceedings{
kang2024latent,
title={Latent Paraphrasing: Perturbation on Layers Improves Knowledge Injection in Language Models},
author={Minki Kang and Sung Ju Hwang and Gibbeum Lee and Jaewoong Cho},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T1lFrYwtf7}
} | As Large Language Models (LLMs) are increasingly deployed in specialized domains with continuously evolving knowledge, the need for timely and precise knowledge injection has become essential. Fine-tuning with paraphrased data is a common approach to enhance knowledge injection, yet it faces two significant challenges: high computational costs due to repetitive external model usage and limited sample diversity.
To this end, we introduce LaPael, a latent-level paraphrasing method that applies input-dependent noise to early LLM layers.
This approach enables diverse and semantically consistent augmentations directly within the model. Furthermore, it eliminates the recurring costs of paraphrase generation for each knowledge update.
Our extensive experiments on question-answering benchmarks demonstrate that LaPael improves knowledge injection over standard fine-tuning and existing noise-based approaches.
Additionally, combining LaPael with data-level paraphrasing further enhances performance. | Latent Paraphrasing: Perturbation on Layers Improves Knowledge Injection in Language Models | [
"Minki Kang",
"Sung Ju Hwang",
"Gibbeum Lee",
"Jaewoong Cho"
] | NeurIPS.cc/2024/Conference | 2411.00686 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=T0glCBw28a | @inproceedings{
huang2024the,
title={The {ALCHE}mist: Automated Labeling 500x {CHE}aper than {LLM} Data Annotators},
author={Tzu-Heng Huang and Catherine Cao and Vaishnavi Bhargava and Frederic Sala},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T0glCBw28a}
} | Large pretrained models can be used as annotators, helping replace or augment crowdworkers and enabling distilling generalist models into smaller specialist models. Unfortunately, this comes at a cost: employing top-of-the-line models often requires paying thousands of dollars for API calls, while the resulting datasets are static and challenging to audit. To address these challenges, we propose a simple alternative: rather than directly querying labels from pretrained models, we task models to generate programs that can produce labels. These programs can be stored and applied locally, re-used and extended, and cost orders of magnitude less. Our system, $\textbf{Alchemist}$, obtains comparable to or better performance than large language model-based annotation in a range of tasks for a fraction of the cost: on average, improvements amount to a $\textbf{12.9}$% enhancement while the total labeling costs across all datasets are reduced by a factor of approximately $\textbf{500}\times$. | The ALCHEmist: Automated Labeling 500x CHEaper than LLM Data Annotators | [
"Tzu-Heng Huang",
"Catherine Cao",
"Vaishnavi Bhargava",
"Frederic Sala"
] | NeurIPS.cc/2024/Conference | 2407.11004 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=T0e4Nw09XX | @inproceedings{
hanneke2024universal,
title={Universal Rates for Active Learning},
author={Steve Hanneke and Amin Karbasi and Shay Moran and Grigoris Velegkas},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T0e4Nw09XX}
} | In this work we study the problem of actively learning binary classifiers
from a given concept class, i.e., learning by utilizing unlabeled data
and submitting targeted queries about their labels to a domain expert.
We evaluate the quality of our solutions by considering the learning curves
they induce, i.e., the rate of decrease
of the misclassification probability as the number of label queries
increases. The majority of the literature on active learning has
focused on obtaining uniform guarantees on the error rate which are
only able to explain the upper envelope of the learning curves over families
of different data-generating distributions. We diverge from this line of
work and we focus on the distribution-dependent framework of universal
learning whose goal is to obtain guarantees that hold for any fixed distribution,
but do not apply uniformly over all the distributions. We provide a
complete characterization of the optimal learning rates that are achievable
by algorithms that have to specify the number of unlabeled examples they
use ahead of their execution. Moreover, we identify combinatorial complexity
measures that give rise to each case of our tetrachotomic characterization.
This resolves an open question that was posed by Balcan et al. (2010).
As a byproduct of our main result,
we develop an active learning algorithm for partial concept classes
that achieves exponential learning rates in the uniform setting. | Universal Rates for Active Learning | [
"Steve Hanneke",
"Amin Karbasi",
"Shay Moran",
"Grigoris Velegkas"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=T0axIflVDD | @inproceedings{
ye2024frequency,
title={Frequency Adaptive Normalization For Non-stationary Time Series Forecasting},
author={Weiwei Ye and Songgaojun Deng and Qiaosha Zou and Ning Gui},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T0axIflVDD}
} | Time series forecasting typically needs to address non-stationary data with evolving trend and seasonal patterns. To address the non-stationarity, reversible instance normalization has been recently proposed to alleviate impacts from the trend with certain statistical measures, e.g., mean and variance. Although they demonstrate improved predictive accuracy, they are limited to expressing basic trends and are incapable of handling seasonal patterns. To address this limitation, this paper proposes a new instance normalization solution, called frequency adaptive normalization (FAN), which extends instance normalization in handling both dynamic trend and seasonal patterns. Specifically, we employ the Fourier transform to identify instance-wise predominant frequent components that cover most non-stationary factors.
Furthermore, the discrepancy of those frequency components between inputs and outputs is explicitly modeled as a prediction task with a simple MLP model. FAN is a model-agnostic method that can be applied to arbitrary predictive backbones. We instantiate FAN on four widely used forecasting models as the backbone and evaluate their prediction performance improvements on eight benchmark datasets. FAN demonstrates significant performance advancement, achieving 7.76\%$\sim$37.90\% average improvements in MSE. Our code is publicly available at http://github.com/icannotnamemyself/FAN. | Frequency Adaptive Normalization For Non-stationary Time Series Forecasting | [
"Weiwei Ye",
"Songgaojun Deng",
"Qiaosha Zou",
"Ning Gui"
] | NeurIPS.cc/2024/Conference | 2409.20371 | [
"https://github.com/wayne155/FAN"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=T07OHxcEYP | @inproceedings{
qiao2024differentially,
title={Differentially Private Reinforcement Learning with Self-Play},
author={Dan Qiao and Yu-Xiang Wang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=T07OHxcEYP}
} | We study the problem of multi-agent reinforcement learning (multi-agent RL) with differential privacy (DP) constraints. This is well-motivated by various real-world applications involving sensitive data, where it is critical to protect users' private information. We first extend the definitions of Joint DP (JDP) and Local DP (LDP) to two-player zero-sum episodic Markov Games, where both definitions ensure trajectory-wise privacy protection. Then we design a provably efficient algorithm based on optimistic Nash value iteration and privatization of Bernstein-type bonuses. The algorithm is able to satisfy JDP and LDP requirements when instantiated with appropriate privacy mechanisms. Furthermore, for both notions of DP, our regret bound generalizes the best known result under the single-agent RL case, while our regret could also reduce to the best known result for multi-agent RL without privacy constraints. To the best of our knowledge, these are the first results towards understanding trajectory-wise privacy protection in multi-agent RL. | Differentially Private Reinforcement Learning with Self-Play | [
"Dan Qiao",
"Yu-Xiang Wang"
] | NeurIPS.cc/2024/Conference | 2404.07559 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SyMhGilvCv | @inproceedings{
jain2024prompt,
title={Prompt Tuning Strikes Back: Customizing Foundation Models with Low-Rank Prompt Adaptation},
author={Abhinav Jain and Swarat Chaudhuri and Thomas Reps and Chris Jermaine},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SyMhGilvCv}
} | Parameter-Efficient Fine-Tuning (PEFT) has become the standard for customising Foundation Models (FMs) to user-specific downstream tasks. However, typical PEFT methods require storing multiple task-specific adapters, creating scalability issues as these adapters must be housed and run at the FM server. Traditional prompt tuning offers a potential solution by customising them through task-specific input prefixes, but it under-performs compared to other PEFT methods like LoRA. To address this gap, we propose Low-Rank Prompt Adaptation (LoPA), a prompt-tuning-based approach that performs on par with state-of-the-art PEFT methods and full fine-tuning while being more parameter-efficient and not requiring a server-based adapter. LoPA generates soft prompts by balancing between sharing task-specific information across instances and customization for each instance. It uses a low-rank decomposition of the soft-prompt component encoded for each instance to achieve parameter efficiency. We provide a comprehensive evaluation on multiple natural language understanding and code generation and understanding tasks across a wide range of foundation models with varying sizes. | Prompt Tuning Strikes Back: Customizing Foundation Models with Low-Rank Prompt Adaptation | [
"Abhinav Jain",
"Swarat Chaudhuri",
"Thomas Reps",
"Chris Jermaine"
] | NeurIPS.cc/2024/Conference | 2405.15282 | [
"https://github.com/jabhinav/prompt-tuning-strikes-back-with-lopa"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SxRblm9aMs | @inproceedings{
yau2024are,
title={Are Graph Neural Networks Optimal Approximation Algorithms?},
author={Morris Yau and Nikolaos Karalias and Eric Hanqing Lu and Jessica Xu and Stefanie Jegelka},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SxRblm9aMs}
} | In this work we design graph neural network architectures that capture optimal
approximation algorithms for a large class of combinatorial optimization problems,
using powerful algorithmic tools from semidefinite programming (SDP). Concretely, we prove that polynomial-sized message-passing algorithms can represent
the most powerful polynomial time algorithms for Max Constraint Satisfaction
Problems assuming the Unique Games Conjecture. We leverage this result to
construct efficient graph neural network architectures, OptGNN, that obtain high quality approximate solutions on landmark combinatorial optimization problems
such as Max-Cut, Min-Vertex-Cover, and Max-3-SAT. Our approach achieves
strong empirical results across a wide range of real-world and synthetic datasets
against solvers and neural baselines. Finally, we take advantage of OptGNN’s
ability to capture convex relaxations to design an algorithm for producing bounds
on the optimal solution from the learned embeddings of OptGNN. | Are Graph Neural Networks Optimal Approximation Algorithms? | [
"Morris Yau",
"Nikolaos Karalias",
"Eric Hanqing Lu",
"Jessica Xu",
"Stefanie Jegelka"
] | NeurIPS.cc/2024/Conference | 2310.00526 | [
"https://github.com/penlu/bespoke-gnn4do"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=Swh8LxuycA | @inproceedings{
nath2024learning,
title={Learning Goal-Conditioned Representations for Language Reward Models},
author={Vaskar Nath and Dylan Z Slack and Jeff Da and Yuntao Ma and Hugh Zhang and Spencer Whitehead and Sean M. Hendryx},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=Swh8LxuycA}
} | Techniques that learn improved representations via offline data or self-supervised objectives have shown impressive results in traditional reinforcement learning.
Nevertheless, it is unclear how improved representation learning can benefit reinforcement learning from human feedback on language models.
In this work, we propose training reward models (RMs) in a contrastive, $\textit{goal-conditioned}$ fashion by increasing the representation similarity of future states along sampled preferred trajectories and decreasing the similarity along randomly sampled dispreferred trajectories.
This objective significantly improves reward model performance by up to 0.09 AUROC across challenging benchmarks, such as MATH and GSM8k. These findings extend to general alignment as well -- on the Helpful-Harmless dataset, we observe 2.3\% increase in accuracy.
Beyond improving reward model performance, we show this way of training RM representations enables improved steerability because it allows us to evaluate the likelihood of an action achieving a particular goal-state (e.g. whether a solution is correct or helpful).
Leveraging this insight, we find that we can filter up to 55\% of generated tokens during majority voting by discarding trajectories likely to end up in an "incorrect" state, which leads to significant cost savings.
We additionally find that these representations can perform fine-grained control by conditioning on desired future goal-states.
For example, we show that steering a Llama 3 model towards helpful generations with our approach improves helpfulness by $9.6$\% over a supervised-fine-tuning trained baseline.
Similarly, steering the model towards complex generations improves complexity by $21.6$\% over the baseline.
Overall, we find that training RMs in this contrastive, goal-conditioned fashion significantly improves performance and enables model steerability. | Learning Goal-Conditioned Representations for Language Reward Models | [
"Vaskar Nath",
"Dylan Z Slack",
"Jeff Da",
"Yuntao Ma",
"Hugh Zhang",
"Spencer Whitehead",
"Sean M. Hendryx"
] | NeurIPS.cc/2024/Conference | 2407.13887 | [
"https://github.com/vaskarnathscale/goal-conditioned-rm"
] | https://huggingface.co/papers/2407.13887 | 1 | 0 | 0 | 7 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=SvmJJJS0q1 | @inproceedings{
reddy2024detecting,
title={Detecting and Measuring Confounding Using Causal Mechanism Shifts},
author={Abbavaram Gowtham Reddy and Vineeth N. Balasubramanian},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SvmJJJS0q1}
} | Detecting and measuring confounding effects from data is a key challenge in causal inference. Existing methods frequently assume causal sufficiency, disregarding the presence of unobserved confounding variables. Causal sufficiency is both unrealistic and empirically untestable. Additionally, existing methods make strong parametric assumptions about the underlying causal generative process to guarantee the identifiability of confounding variables. Relaxing the causal sufficiency and parametric assumptions and leveraging recent advancements in causal discovery and confounding analysis with non-i.i.d. data, we propose a comprehensive approach for detecting and measuring confounding. We consider various definitions of confounding and introduce tailored methodologies to achieve three objectives: (i) detecting and measuring confounding among a set of variables, (ii) separating observed and unobserved confounding effects, and (iii) understanding the relative strengths of confounding bias between different sets of variables. We present useful properties of a confounding measure and present measures that satisfy those properties. Our empirical results support the usefulness of the proposed measures. | Detecting and Measuring Confounding Using Causal Mechanism Shifts | [
"Abbavaram Gowtham Reddy",
"Vineeth N. Balasubramanian"
] | NeurIPS.cc/2024/Conference | 2409.17840 | [
"https://github.com/gautam0707/cd_cnf"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SuLxkxCENa | @inproceedings{
georgiev2024deep,
title={Deep Equilibrium Algorithmic Reasoning},
author={Dobrik Georgiev Georgiev and JJ Wilson and Davide Buffelli and Pietro Lio},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SuLxkxCENa}
} | Neural Algorithmic Reasoning (NAR) research has demonstrated that graph neural networks (GNNs) could learn to execute classical algorithms. However, most previous approaches have always used a recurrent architecture, where each iteration of the GNN matches an iteration of the algorithm. In this paper we study neurally solving algorithms from a different perspective: since the algorithm’s solution is often an equilibrium, it is possible to find the solution directly by solving an equilibrium equation. Our approach requires no information on the ground-truth number of steps of the algorithm, both during train and test time. Furthermore, the proposed method improves the performance of GNNs on executing algorithms and is a step towards speeding up existing NAR models. Our empirical evidence, leveraging algorithms from the CLRS-30 benchmark, validates that one can train a network to solve algorithmic problems by directly finding the equilibrium. We discuss the practical implementation of such models and propose regularisations to improve the performance of these equilibrium reasoners. | Deep Equilibrium Algorithmic Reasoning | [
"Dobrik Georgiev Georgiev",
"JJ Wilson",
"Davide Buffelli",
"Pietro Lio"
] | NeurIPS.cc/2024/Conference | 2410.15059 | [
"https://github.com/HekpoMaH/DEAR"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=StapcUWm9q | @inproceedings{
yang2024diffusion,
title={Diffusion Model with Cross Attention as an Inductive Bias for Disentanglement},
author={Tao Yang and Cuiling Lan and Yan Lu and Nanning Zheng},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=StapcUWm9q}
} | Disentangled representation learning strives to extract the intrinsic factors within observed data. Factorizing these representations in an unsupervised manner is notably challenging and usually requires tailored loss functions or specific structural designs. In this paper, we introduce a new perspective and framework, demonstrating that diffusion models with cross-attention can serve as a powerful inductive bias to facilitate the learning of disentangled representations. We propose to encode an image to a set of concept tokens and treat them as the condition of the latent diffusion for image reconstruction, where cross-attention over the concept tokens is used to bridge the interaction between the encoder and diffusion. Without any additional regularization, this framework achieves superior disentanglement performance on the benchmark datasets, surpassing all previous methods with intricate designs. We have conducted comprehensive ablation studies and visualization analysis, shedding light on the functioning of this model. We anticipate that our findings will inspire more investigation on exploring diffusion for disentangled representation learning towards more sophisticated data analysis and understanding. | Diffusion Model with Cross Attention as an Inductive Bias for Disentanglement | [
"Tao Yang",
"Cuiling Lan",
"Yan Lu",
"Nanning Zheng"
] | NeurIPS.cc/2024/Conference | 2402.09712 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=Ss7l98DVvD | @inproceedings{
xu2024wildgs,
title={Wild-{GS}: Real-Time Novel View Synthesis from Unconstrained Photo Collections},
author={Jiacong Xu and Yiqun Mei and Vishal M. Patel},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=Ss7l98DVvD}
} | Photographs captured in unstructured tourist environments frequently exhibit variable appearances and transient occlusions, challenging accurate scene reconstruction and inducing artifacts in novel view synthesis. Although prior approaches have integrated the Neural Radiance Field (NeRF) with additional learnable modules to handle the dynamic appearances and eliminate transient objects, their extensive training demands and slow rendering speeds limit practical deployments. Recently, 3D Gaussian Splatting (3DGS) has emerged as a promising alternative to NeRF, offering superior training and inference efficiency along with better rendering quality. This paper presents \textit{Wild-GS}, an innovative adaptation of 3DGS optimized for unconstrained photo collections while preserving its efficiency benefits. \textit{Wild-GS} determines the appearance of each 3D Gaussian by their inherent material attributes, global illumination and camera properties per image, and point-level local variance of reflectance. Unlike previous methods that model reference features in image space, \textit{Wild-GS} explicitly aligns the pixel appearance features to the corresponding local Gaussians by sampling the triplane extracted from the reference image. This novel design effectively transfers the high-frequency detailed appearance of the reference view to 3D space and significantly expedites the training process. Furthermore, 2D visibility maps and depth regularization are leveraged to mitigate the transient effects and constrain the geometry, respectively. Extensive experiments demonstrate that \textit{Wild-GS} achieves state-of-the-art rendering performance and the highest efficiency in both training and inference among all the existing techniques. The code can be accessed via: https://github.com/XuJiacong/Wild-GS | Wild-GS: Real-Time Novel View Synthesis from Unconstrained Photo Collections | [
"Jiacong Xu",
"Yiqun Mei",
"Vishal M. Patel"
] | NeurIPS.cc/2024/Conference | 2406.10373 | [
""
] | https://huggingface.co/papers/2406.10373 | 0 | 0 | 0 | 3 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=SrQua0ATRZ | @inproceedings{
wang2024diffusioninspired,
title={Diffusion-Inspired Truncated Sampler for Text-Video Retrieval},
author={Jiamian Wang and Pichao WANG and Dongfang Liu and Qiang Guan and Sohail Dianat and MAJID RABBANI and Raghuveer Rao and ZHIQIANG TAO},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SrQua0ATRZ}
} | Prevalent text-to-video retrieval methods represent multimodal text-video data in a joint embedding space, aiming at bridging the relevant text-video pairs and pulling away irrelevant ones. One main challenge in state-of-the-art retrieval methods lies in the modality gap, which stems from the substantial disparities between text and video and can persist in the joint space. In this work, we leverage the potential of Diffusion models to address the text-video modality gap by progressively aligning text and video embeddings in a unified space. However, we identify two key limitations of existing Diffusion models in retrieval tasks: The L2 loss does not fit the ranking problem inherent in text-video retrieval, and the generation quality heavily depends on the varied initial point drawn from the isotropic Gaussian, causing inaccurate retrieval. To this end, we introduce a new Diffusion-Inspired Truncated Sampler (DITS) that jointly performs progressive alignment and modality gap modeling in the joint embedding space. The key innovation of DITS is to leverage the inherent proximity of text and video embeddings, defining a truncated diffusion flow from the fixed text embedding to the video embedding, enhancing controllability compared to adopting the isotropic Gaussian. Moreover, DITS adopts the contrastive loss to jointly consider the relevant and irrelevant pairs, not only facilitating alignment but also yielding a discriminatively structured embedding. Experiments on five benchmark datasets suggest the state-of-the-art performance of DITS. We empirically find that DITS can also improve the structure of the CLIP embedding space. Code is available at https://github.com/Jiamian- Wang/DITS-text-video-retrieval | Diffusion-Inspired Truncated Sampler for Text-Video Retrieval | [
"Jiamian Wang",
"Pichao WANG",
"Dongfang Liu",
"Qiang Guan",
"Sohail Dianat",
"MAJID RABBANI",
"Raghuveer Rao",
"ZHIQIANG TAO"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SrFbgIjb53 | @inproceedings{
du2024mogu,
title={Mo{GU}: A Framework for Enhancing Safety of {LLM}s While Preserving Their Usability},
author={Yanrui Du and Sendong Zhao and Danyang Zhao and Ming Ma and Yuhan Chen and Liangyu Huo and Qing Yang and Dongliang Xu and Bing Qin},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SrFbgIjb53}
} | Large Language Models (LLMs) are increasingly deployed in various applications. As their usage grows, concerns regarding their safety are rising, especially in maintaining harmless responses when faced with malicious instructions. Many defense strategies have been developed to enhance the safety of LLMs. However, our research finds that existing defense strategies lead LLMs to predominantly adopt a rejection-oriented stance, thereby diminishing the usability of their responses to benign instructions. To solve this problem, we introduce the MoGU framework, designed to enhance LLMs' safety while preserving their usability. Our MoGU framework transforms the base LLM into two variants: the usable LLM and the safe LLM, and further employs dynamic routing to balance their contribution. When encountering malicious instructions, the router will assign a higher weight to the safe LLM to ensure that responses are harmless. Conversely, for benign instructions, the router prioritizes the usable LLM, facilitating usable and helpful responses. On various open-sourced LLMs, we compare multiple defense strategies to verify the superiority of our MoGU framework. Besides, our analysis provides key insights into the effectiveness of MoGU and verifies that our designed routing mechanism can effectively balance the contribution of each variant by assigning weights. Our work released the safer Llama2, Vicuna, Falcon, Dolphin, and Baichuan2. | MoGU: A Framework for Enhancing Safety of LLMs While Preserving Their Usability | [
"Yanrui Du",
"Sendong Zhao",
"Danyang Zhao",
"Ming Ma",
"Yuhan Chen",
"Liangyu Huo",
"Qing Yang",
"Dongliang Xu",
"Bing Qin"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SpcEwP6EYt | @inproceedings{
guo2024enofsnn,
title={En{OF}-{SNN}: Training Accurate Spiking Neural Networks via Enhancing the Output Feature},
author={Yufei Guo and Weihang Peng and Xiaode Liu and Yuanpei Chen and Yuhan Zhang and Xin Tong and Zhou Jie and Zhe Ma},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SpcEwP6EYt}
} | Spiking neural networks (SNNs) have gained more and more interest as one of the energy-efficient alternatives of conventional artificial neural networks (ANNs). They exchange 0/1 spikes for processing information, thus most of the multiplications in networks can be replaced by additions. However, binary spike feature maps will limit the expressiveness of the SNN and result in unsatisfactory performance compared with ANNs.
It is shown that a rich output feature representation, i.e., the feature vector before classifier) is beneficial to training an accurate model in ANNs for classification.
We wonder if it also does for SNNs and how to improve the feature representation of the SNN.
To this end, we materialize this idea in two special designed methods for SNNs.
First, inspired by some ANN-SNN methods that directly copy-paste the weight parameters from trained ANN with light modification to homogeneous SNN can obtain a well-performed SNN, we use rich information of the weight parameters from the trained ANN counterpart to guide the feature representation learning of the SNN.
In particular, we present the SNN's and ANN's feature representation from the same input to ANN's classifier to product SNN's and ANN's outputs respectively and then align the feature with the KL-divergence loss as in knowledge distillation methods, called L_ AF loss.
It can be seen as a novel and effective knowledge distillation method specially designed for the SNN that comes from both the knowledge distillation and ANN-SNN methods. Various ablation study shows that the L_AF loss is more powerful than the vanilla knowledge distillation method.
Second, we replace the last Leaky Integrate-and-Fire (LIF) activation layer as the ReLU activation layer to generate the output feature, thus a more powerful SNN with full-precision feature representation can be achieved but with only a little extra computation.
Experimental results show that our method consistently outperforms the current state-of-the-art algorithms on both popular non-spiking static and neuromorphic datasets. We provide an extremely simple but effective way to train high-accuracy spiking neural networks. | EnOF-SNN: Training Accurate Spiking Neural Networks via Enhancing the Output Feature | [
"Yufei Guo",
"Weihang Peng",
"Xiaode Liu",
"Yuanpei Chen",
"Yuhan Zhang",
"Xin Tong",
"Zhou Jie",
"Zhe Ma"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SpPAB1tmlC | @inproceedings{
diao2024unveiling,
title={Unveiling Encoder-Free Vision-Language Models},
author={Haiwen Diao and Yufeng Cui and Xiaotong Li and Yueze Wang and Huchuan Lu and Xinlong Wang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SpPAB1tmlC}
} | Existing vision-language models (VLMs) mostly rely on vision encoders to extract visual features followed by large language models (LLMs) for visual-language tasks. However, the vision encoders set a strong inductive bias in abstracting visual representation, e.g., resolution, aspect ratio, and semantic priors, which could impede the flexibility and efficiency of the VLMs. Training pure VLMs that accept the seamless vision and language inputs, i.e., without vision encoders, remains challenging and rarely explored. Empirical observations reveal that direct training without encoders results in slow convergence and large performance gaps. In this work, we bridge the gap between encoder-based and encoder-free models, and present a simple yet effective training recipe towards pure VLMs. Specifically, we unveil the key aspects of training encoder-free VLMs efficiently via thorough experiments: (1) Bridging vision-language representation inside one unified decoder; (2) Enhancing visual recognition capability via extra supervision. With these strategies, we launch EVE, an encoder-free vision-language model that can be trained and forwarded efficiently. Notably, solely utilizing 35M publicly accessible data, EVE can impressively rival the encoder-based VLMs of similar capacities across multiple vision-language benchmarks. It significantly outperforms the counterpart Fuyu-8B with mysterious training procedures and undisclosed training data. We believe that EVE provides a transparent and efficient route for developing pure decoder-only architecture across modalities. | Unveiling Encoder-Free Vision-Language Models | [
"Haiwen Diao",
"Yufeng Cui",
"Xiaotong Li",
"Yueze Wang",
"Huchuan Lu",
"Xinlong Wang"
] | NeurIPS.cc/2024/Conference | 2406.11832 | [
"https://github.com/baaivision/eve"
] | https://huggingface.co/papers/2406.11832 | 4 | 49 | 3 | 6 | [
"BAAI/EVE-7B-HD-v1.0",
"BAAI/EVE-7B-v1.0",
"BAAI/EVE-7B-Pretrain-v1.0"
] | [] | [] | [
"BAAI/EVE-7B-HD-v1.0",
"BAAI/EVE-7B-v1.0",
"BAAI/EVE-7B-Pretrain-v1.0"
] | [] | [] | 1 | oral |
null | https://openreview.net/forum?id=SoYCqMiVIh | @inproceedings{
wijeratne2024unscrambling,
title={Unscrambling disease progression at scale: fast inference of event permutations with optimal transport},
author={Peter A. Wijeratne and Daniel C. Alexander},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SoYCqMiVIh}
} | Disease progression models infer group-level temporal trajectories of change in patients' features as a chronic degenerative condition plays out. They provide unique insight into disease biology and staging systems with individual-level clinical utility. Discrete models consider disease progression as a latent permutation of events, where each event corresponds to a feature becoming measurably abnormal. However, permutation inference using traditional maximum likelihood approaches becomes prohibitive due to combinatoric explosion, severely limiting model dimensionality and utility. Here we leverage ideas from optimal transport to model disease progression as a latent permutation matrix of events belonging to the Birkhoff polytope, facilitating fast inference via optimisation of the variational lower bound. This enables a factor of 1000 times faster inference than the current state of the art and, correspondingly, supports models with several orders of magnitude more features than the current state of the art can consider. Experiments demonstrate the increase in speed, accuracy and robustness to noise in simulation. Further experiments with real-world imaging data from two separate datasets, one from Alzheimer's disease patients, the other age-related macular degeneration, showcase, for the first time, pixel-level disease progression events in the brain and eye, respectively. Our method is low compute, interpretable and applicable to any progressive condition and data modality, giving it broad potential clinical utility. | Unscrambling disease progression at scale: fast inference of event permutations with optimal transport | [
"Peter A. Wijeratne",
"Daniel C. Alexander"
] | NeurIPS.cc/2024/Conference | 2410.14388 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SoTK84ewb7 | @inproceedings{
zhou2024zeroshot,
title={Zero-Shot Scene Reconstruction from Single Images with Deep Prior Assembly},
author={Junsheng Zhou and Yu-Shen Liu and Zhizhong Han},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SoTK84ewb7}
} | Large language and vision models have been leading a revolution in visual computing. By greatly scaling up sizes of data and model parameters, the large models learn deep priors which lead to remarkable performance in various tasks. In this work, we present deep prior assembly, a novel framework that assembles diverse deep priors from large models for scene reconstruction from single images in a zero-shot manner. We show that this challenging task can be done without extra knowledge but just simply generalizing one deep prior in one sub-task. To this end, we introduce novel methods related to poses, scales, and occlusion parsing which are keys to enable deep priors to work together in a robust way. Deep prior assembly does not require any 3D or 2D data-driven training in the task and demonstrates superior performance in generalizing priors to open-world scenes. We conduct evaluations on various datasets, and report analysis, numerical and visual comparisons with the latest methods to show our superiority. Project page: https://junshengzhou.github.io/DeepPriorAssembly. | Zero-Shot Scene Reconstruction from Single Images with Deep Prior Assembly | [
"Junsheng Zhou",
"Yu-Shen Liu",
"Zhizhong Han"
] | NeurIPS.cc/2024/Conference | 2410.15971 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SoM3vngOH5 | @inproceedings{
mehrotra2024tree,
title={Tree of Attacks: Jailbreaking Black-Box {LLM}s Automatically},
author={Anay Mehrotra and Manolis Zampetakis and Paul Kassianik and Blaine Nelson and Hyrum S Anderson and Yaron Singer and Amin Karbasi},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SoM3vngOH5}
} | While Large Language Models (LLMs) display versatile functionality, they continue to generate harmful, biased, and toxic content, as demonstrated by the prevalence of human-designed *jailbreaks*. In this work, we present *Tree of Attacks with Pruning* (TAP), an automated method for generating jailbreaks that only requires black-box access to the target LLM. TAP utilizes an attacker LLM to iteratively refine candidate (attack) prompts until one of the refined prompts jailbreaks the target. In addition, before sending prompts to the target, TAP assesses them and prunes the ones unlikely to result in jailbreaks, reducing the number of queries sent to the target LLM. In empirical evaluations, we observe that TAP generates prompts that jailbreak state-of-the-art LLMs (including GPT4-Turbo and GPT4o) for more than 80% of the prompts. This significantly improves upon the previous state-of-the-art black-box methods for generating jailbreaks while using a smaller number of queries than them. Furthermore, TAP is also capable of jailbreaking LLMs protected by state-of-the-art *guardrails*, e.g., LlamaGuard. | Tree of Attacks: Jailbreaking Black-Box LLMs Automatically | [
"Anay Mehrotra",
"Manolis Zampetakis",
"Paul Kassianik",
"Blaine Nelson",
"Hyrum S Anderson",
"Yaron Singer",
"Amin Karbasi"
] | NeurIPS.cc/2024/Conference | 2312.02119 | [
"https://github.com/ricommunity/tap"
] | https://huggingface.co/papers/2312.02119 | 0 | 1 | 0 | 7 | [] | [] | [
"latticeflow/compl-ai-board",
"TrustSafeAI/GradientCuff-Jailbreak-Defense",
"TrustSafeAI/Defensive-Prompt-Patch-Jailbreak-Defense"
] | [] | [] | [
"latticeflow/compl-ai-board",
"TrustSafeAI/GradientCuff-Jailbreak-Defense",
"TrustSafeAI/Defensive-Prompt-Patch-Jailbreak-Defense"
] | 1 | poster |
null | https://openreview.net/forum?id=SnTxbQSrW7 | @inproceedings{
li2024adapting,
title={Adapting to Unknown Low-Dimensional Structures in Score-Based Diffusion Models},
author={Gen Li and Yuling Yan},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SnTxbQSrW7}
} | This paper investigates score-based diffusion models when the underlying target distribution is concentrated on or near low-dimensional manifolds within the higher-dimensional space in which they formally reside, a common characteristic of natural image distributions. Despite previous efforts to understand the data generation process of diffusion models, existing theoretical support remains highly suboptimal in the presence of low-dimensional structure, which we strengthen in this paper. For the popular Denoising Diffusion Probabilistic Model (DDPM), we find that the dependency of the error incurred within each denoising step on the ambient dimension $d$ is in general unavoidable. We further identify a unique design of coefficients that yields a converges rate at the order of $O(k^{2}/\sqrt{T})$ (up to log factors), where $k$ is the intrinsic dimension of the target distribution and $T$ is the number of steps. This represents the first theoretical demonstration that the DDPM sampler can adapt to unknown low-dimensional structures in the target distribution, highlighting the critical importance of coefficient design. All of this is achieved by a novel set of analysis tools that characterize the algorithmic dynamics in a more deterministic manner. | Adapting to Unknown Low-Dimensional Structures in Score-Based Diffusion Models | [
"Gen Li",
"Yuling Yan"
] | NeurIPS.cc/2024/Conference | 2405.14861 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SlDx451MjC | @inproceedings{
bilecen2024dual,
title={Dual Encoder {GAN} Inversion for High-Fidelity 3D Head Reconstruction from Single Images},
author={Bahri Batuhan Bilecen and Ahmet Berke G{\"o}kmen and Aysegul Dundar},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SlDx451MjC}
} | 3D GAN inversion aims to project a single image into the latent space of a 3D Generative Adversarial Network (GAN), thereby achieving 3D geometry reconstruction. While there exist encoders that achieve good results in 3D GAN inversion, they are predominantly built on EG3D, which specializes in synthesizing near-frontal views and is limiting in synthesizing comprehensive 3D scenes from diverse viewpoints. In contrast to existing approaches, we propose a novel framework built on PanoHead, which excels in synthesizing images from a 360-degree perspective. To achieve realistic 3D modeling of the input image, we introduce a dual encoder system tailored for high-fidelity reconstruction and realistic generation from different viewpoints. Accompanying this, we propose a stitching framework on the triplane domain to get the best predictions from both. To achieve seamless stitching, both encoders must output consistent results despite being specialized for different tasks. For this reason, we carefully train these encoders using specialized losses, including an adversarial loss based on our novel occlusion-aware triplane discriminator. Experiments reveal that our approach surpasses the existing encoder training methods qualitatively and quantitatively. | Dual Encoder GAN Inversion for High-Fidelity 3D Head Reconstruction from Single Images | [
"Bahri Batuhan Bilecen",
"Ahmet Berke Gökmen",
"Aysegul Dundar"
] | NeurIPS.cc/2024/Conference | 2409.20530 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=Skv26JteFz | @inproceedings{
aliakbarpour2024optimal,
title={Optimal Hypothesis Selection in (Almost) Linear Time},
author={Maryam Aliakbarpour and Mark Bun and Adam Smith},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=Skv26JteFz}
} | Hypothesis selection, also known as density estimation, is a fundamental problem in statistics and learning theory. Suppose we are given a sample set from an unknown distribution $P$ and a finite class of candidate distributions (called hypotheses) $\mathcal{H} \coloneqq \{H_1, H_2, \ldots, H_n\}$. The aim is to design an algorithm that selects a distribution $\hat H$ in $\mathcal{H}$ that best fits the data. The algorithm's accuracy is measured based on the distance between $\hat{H}$ and $P$ compared to the distance of the closest distribution in $\mathcal{H}$ to $P$ (denoted by $OPT$). Concretely, we aim for $\|\hat{H} - P\|_{TV}$ to be at most $ \alpha \cdot OPT + \epsilon$ for some small $\epsilon$ and $\alpha$.
While it is possible to decrease the value of $\epsilon$ as the number of samples increases, $\alpha$ is an inherent characteristic of the algorithm. In fact, one cannot hope to achieve $\alpha < 3$ even when there are only two candidate hypotheses, unless the number of samples is proportional to the domain size of $P$ [Bousquet, Kane, Moran '19]. Finding the best $\alpha$ has been one of the main focuses of studies of the problem since early work of [Devroye, Lugosi '01]. Prior to our work, no algorithm was known that achieves $\alpha = 3$ in near-linear time. We provide the first algorithm that operates in almost linear time ($\tilde{O}(n/\epsilon^3)$ time) and achieves $\alpha = 3$. This result improves upon a long list of results in hypothesis selection. Previously known algorithms either had worse time complexity, a larger factor $\alpha$, or extra assumptions about the problem setting.
In addition to this algorithm, we provide another (almost) linear-time algorithm with better dependency on the additive accuracy parameter $\epsilon$, albeit with a slightly worse accuracy parameter, $\alpha = 4$. | Optimal Hypothesis Selection in (Almost) Linear Time | [
"Maryam Aliakbarpour",
"Mark Bun",
"Adam Smith"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=Sk2duBGvrK | @inproceedings{
li2024understanding,
title={Understanding Generalizability of Diffusion Models Requires Rethinking the Hidden Gaussian Structure},
author={Xiang Li and Yixiang Dai and Qing Qu},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=Sk2duBGvrK}
} | In this work, we study the generalizability of diffusion models by looking into the hidden properties of the learned score functions, which are essentially a series of deep denoisers trained on various noise levels. We observe that as diffusion models transition from memorization to generalization, their corresponding nonlinear diffusion denoisers exhibit increasing linearity. This discovery leads us to investigate the linear counterparts of the nonlinear diffusion models, which are a series of linear models trained to match the function mappings of the nonlinear diffusion denoisers. Surprisingly, these linear denoisers are approximately the optimal denoisers for a multivariate Gaussian distribution characterized by the empirical mean and covariance of the training dataset. This finding implies that diffusion models have the inductive bias towards capturing and utilizing the Gaussian structure (covariance information) of the training dataset for data generation. We empirically demonstrate that this inductive bias is a unique property of diffusion models in the generalization regime, which becomes increasingly evident when the model's capacity is relatively small compared to the training dataset size. In the case that the model is highly overparameterized, this inductive bias emerges during the initial training phases before the model fully memorizes its training data. Our study provides crucial insights into understanding the notable strong generalization phenomenon recently observed in real-world diffusion models. | Understanding Generalizability of Diffusion Models Requires Rethinking the Hidden Gaussian Structure | [
"Xiang Li",
"Yixiang Dai",
"Qing Qu"
] | NeurIPS.cc/2024/Conference | 2410.24060 | [
"https://github.com/Morefre/Understanding-Generalizability-of-Diffusion-Models-Requires-Rethinking-the-Hidden-Gaussian-Structure"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SjQ1iIqpfU | @inproceedings{
hashemi2024cobo,
title={CoBo: Collaborative Learning via Bilevel Optimization},
author={Diba Hashemi and Lie He and Martin Jaggi},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SjQ1iIqpfU}
} | Collaborative learning is an important tool to train multiple clients more effectively by enabling communication among clients. Identifying helpful clients, however, presents challenging and often introduces significant overhead. In this paper, we model **client-selection** and **model-training** as two interconnected optimization problems, proposing a novel bilevel optimization problem for collaborative learning.
We introduce **CoBo**, a *scalable* and *elastic*, SGD-type alternating optimization algorithm that efficiently addresses these problem with theoretical convergence guarantees. Empirically, **CoBo** achieves superior performance, surpassing popular personalization algorithms by 9.3% in accuracy on a task with high heterogeneity, involving datasets distributed among 80 clients. | CoBo: Collaborative Learning via Bilevel Optimization | [
"Diba Hashemi",
"Lie He",
"Martin Jaggi"
] | NeurIPS.cc/2024/Conference | 2409.05539 | [
"https://github.com/epfml/cobo"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=Sj8G020ADl | @inproceedings{
liao2024inexact,
title={Inexact Augmented Lagrangian Methods for Conic Optimization: Quadratic Growth and Linear Convergence},
author={Feng-Yi Liao and Lijun Ding and Yang Zheng},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=Sj8G020ADl}
} | Augmented Lagrangian Methods (ALMs) are widely employed in solving constrained optimizations, and some efficient solvers are developed based on this framework. Under the quadratic growth assumption, it is known that the dual iterates and the Karush–Kuhn–Tucker (KKT) residuals of ALMs applied to conic programs converge linearly. In contrast, the convergence rate of the primal iterates has remained elusive. In this paper, we resolve this challenge by establishing new $\textit{quadratic growth}$ and $\textit{error bound}$ properties for primal and dual conic programs under the standard strict complementarity condition. Our main results reveal that both primal and dual iterates of the ALMs converge linearly contingent solely upon the assumption of strict complementarity and a bounded solution set. This finding provides a positive answer to an open question regarding the asymptotically linear convergence of the primal iterates of ALMs applied to conic optimization. | Inexact Augmented Lagrangian Methods for Conic Optimization: Quadratic Growth and Linear Convergence | [
"Feng-Yi Liao",
"Lijun Ding",
"Yang Zheng"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SiALFXa0NN | @inproceedings{
teuber2024provably,
title={Provably Safe Neural Network Controllers via Differential Dynamic Logic},
author={Samuel Teuber and Stefan Mitsch and Andre Platzer},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SiALFXa0NN}
} | While neural networks (NNs) have a large potential as autonomous controllers for Cyber-Physical Systems, verifying the safety of neural network based control systems (NNCSs) poses significant challenges for the practical use of NNs— especially when safety is needed for unbounded time horizons. One reason for this is the intractability of analyzing NNs, ODEs and hybrid systems. To this end, we introduce VerSAILLE (Verifiably Safe AI via Logically Linked Envelopes): The first general approach that allows reusing control theory literature for NNCS verification. By joining forces, we can exploit the efficiency of NN verification tools while retaining the rigor of differential dynamic logic (dL). Based on a provably safe control envelope in dL, we derive a specification for the NN which is proven with NN verification tools. We show that a proof of the NN’s adherence to the specification is then mirrored by a dL proof on the infinite-time safety of the NNCS.
The NN verification properties resulting from hybrid systems typically contain nonlinear arithmetic over formulas with arbitrary logical structure while efficient NN verification tools merely support linear constraints. To overcome this divide, we present Mosaic: An efficient, sound and complete verification approach for polynomial real arithmetic properties on piece-wise linear NNs. Mosaic partitions complex NN verification queries into simple queries and lifts off-the-shelf linear constraint tools to the nonlinear setting in a completeness-preserving manner by combining approximation with exact reasoning for counterexample regions. In our evaluation we demonstrate the versatility of VerSAILLE and Mosaic: We prove infinite-time safety on the classical Vertical Airborne Collision Avoidance NNCS verification benchmark for some scenarios while (exhaustively) enumerating counterexample regions in unsafe scenarios. We also show that our approach significantly outperforms the State-of-the-Art tools in closed-loop NNV | Provably Safe Neural Network Controllers via Differential Dynamic Logic | [
"Samuel Teuber",
"Stefan Mitsch",
"Andre Platzer"
] | NeurIPS.cc/2024/Conference | 2402.10998 | [
"https://github.com/samysweb/ncubev"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=Shwtw8uV8l | @inproceedings{
hu2024single,
title={Single Image Reflection Separation via Dual-Stream Interactive Transformers},
author={Qiming Hu and Hainuo Wang and Xiaojie Guo},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=Shwtw8uV8l}
} | Despite satisfactory results on ``easy'' cases of single image reflection separation, prior dual-stream methods still suffer from considerable performance degradation when facing complex ones, i.e, the transmission layer is densely entangled with the reflection having a wide distribution of spatial intensity. The main reasons come from the lack of concern on the feature correlation during interaction, and the limited receptive field. To remedy these deficiencies, this paper presents a Dual-Stream Interactive Transformer (DSIT) design. Specifically, we devise a dual-attention interactive structure that embraces a dual-stream self-attention and a layer-aware dual-stream cross-attention mechanism to simultaneously capture intra-layer and inter-layer feature correlations. Meanwhile, the introduction of attention mechanisms can also mitigate the receptive field limitation. We modulate single-stream pre-trained Transformer embeddings with dual-stream convolutional features through cross-architecture interactions to provide richer semantic priors, thereby further relieving the ill-posedness of the problem. Extensive experimental results reveal the merits of the proposed DSIT over other state-of-the-art alternatives. Our code is publicly available at https://github.com/mingcv/DSIT. | Single Image Reflection Separation via Dual-Stream Interactive Transformers | [
"Qiming Hu",
"Hainuo Wang",
"Xiaojie Guo"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=ShJWT0n7kX | @inproceedings{
du2024doobs,
title={Doob's Lagrangian: A Sample-Efficient Variational Approach to Transition Path Sampling},
author={Yuanqi Du and Michael Plainer and Rob Brekelmans and Chenru Duan and Frank Noe and Carla P Gomes and Alan Aspuru-Guzik and Kirill Neklyudov},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=ShJWT0n7kX}
} | Rare event sampling in dynamical systems is a fundamental problem arising in the natural sciences, which poses significant computational challenges due to an exponentially large space of trajectories. For settings where the dynamical system of interest follows a Brownian motion with known drift, the question of conditioning the process to reach a given endpoint or desired rare event is definitively answered by Doob's $h$-transform. However, the naive estimation of this transform is infeasible, as it requires simulating sufficiently many forward trajectories to estimate rare event probabilities. In this work, we propose a variational formulation of Doob's $h$-transform as an optimization problem over trajectories between a given initial point and the desired ending point. To solve this optimization, we propose a simulation-free training objective with a model parameterization that imposes the desired boundary conditions by design. Our approach significantly reduces the search space over trajectories and avoids expensive trajectory simulation and inefficient importance sampling estimators which are required in existing methods. We demonstrate the ability of our method to find feasible transition paths on real-world molecular simulation and protein folding tasks. | Doob's Lagrangian: A Sample-Efficient Variational Approach to Transition Path Sampling | [
"Yuanqi Du",
"Michael Plainer",
"Rob Brekelmans",
"Chenru Duan",
"Frank Noe",
"Carla P Gomes",
"Alan Aspuru-Guzik",
"Kirill Neklyudov"
] | NeurIPS.cc/2024/Conference | 2410.07974 | [
"https://github.com/plainerman/variational-doob"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=SgODU2mx9T | @inproceedings{
zhuang2024timevarying,
title={Time-Varying Lo{RA}: Towards Effective Cross-Domain Fine-Tuning of Diffusion Models},
author={Zhan Zhuang and Yulong Zhang and Xuehao Wang and Jiangang Lu and Ying Wei and Yu Zhang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SgODU2mx9T}
} | Large-scale diffusion models are adept at generating high-fidelity images and facilitating image editing and interpolation. However, they have limitations when tasked with generating images in dynamic, evolving domains. In this paper, we introduce Terra, a novel Time-varying low-rank adapter that offers a fine-tuning framework specifically tailored for domain flow generation. The key innovation of Terra lies in its construction of a continuous parameter manifold through a time variable, with its expressive power analyzed theoretically. This framework not only enables interpolation of image content and style but also offers a generation-based approach to address the domain shift problems in unsupervised domain adaptation and domain generalization. Specifically, Terra transforms images from the source domain to the target domain and generates interpolated domains with various styles to bridge the gap between domains and enhance the model generalization, respectively. We conduct extensive experiments on various benchmark datasets, empirically demonstrate the effectiveness of Terra. Our source code is publicly available on https://github.com/zwebzone/terra. | Time-Varying LoRA: Towards Effective Cross-Domain Fine-Tuning of Diffusion Models | [
"Zhan Zhuang",
"Yulong Zhang",
"Xuehao Wang",
"Jiangang Lu",
"Ying Wei",
"Yu Zhang"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SeefZa7Vmq | @inproceedings{
wang2024unlearnable,
title={Unlearnable 3D Point Clouds: Class-wise Transformation Is All You Need},
author={Xianlong Wang and Minghui Li and Wei Liu and Hangtao Zhang and Shengshan Hu and Yechao Zhang and Ziqi Zhou and Hai Jin},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SeefZa7Vmq}
} | Traditional unlearnable strategies have been proposed to prevent unauthorized users from training on the 2D image data. With more 3D point cloud data containing sensitivity information, unauthorized usage of this new type data has also become a serious concern. To address this, we propose the first integral unlearnable framework for 3D point clouds including two processes: (i) we propose an unlearnable data protection scheme, involving a class-wise setting established by a category-adaptive allocation strategy and multi-transformations assigned to samples; (ii) we propose a data restoration scheme that utilizes class-wise inverse matrix transformation, thus enabling authorized-only training for unlearnable data. This restoration process is a practical issue overlooked in most existing unlearnable literature, i.e., even authorized users struggle to gain knowledge from 3D unlearnable data. Both theoretical and empirical results (including 6 datasets, 16 models, and 2 tasks) demonstrate the effectiveness of our proposed unlearnable framework. Our code is available at https://github.com/CGCL-codes/UnlearnablePC. | Unlearnable 3D Point Clouds: Class-wise Transformation Is All You Need | [
"Xianlong Wang",
"Minghui Li",
"Wei Liu",
"Hangtao Zhang",
"Shengshan Hu",
"Yechao Zhang",
"Ziqi Zhou",
"Hai Jin"
] | NeurIPS.cc/2024/Conference | 2410.03644 | [
"https://github.com/cgcl-codes/unlearnablepc"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SdLOs1FR4h | @inproceedings{
bommakanti2024fugal,
title={{FUGAL}: Feature-fortified Unrestricted Graph Alignment},
author={Aditya Bommakanti and Harshith Reddy Vonteri and Konstantinos Skitsas and Sayan Ranu and Davide Mottin and Panagiotis Karras},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SdLOs1FR4h}
} | The necessity to align two graphs, minimizing a structural distance metric, is prevalent in biology, chemistry, recommender systems, and social network analysis. Due to the problem’s NP-hardness, prevailing graph alignment methods follow a modular and mediated approach, solving the problem by restricting to the domain of intermediary graph representations or products like embeddings, spectra, and graph signals. Restricting the problem to this intermediate space may distort the original problem and are hence predisposed to miss high-quality solutions. In this paper, we propose an unrestricted method, FUGAL, which finds a permutation matrix that maps one graph to another by directly operating on their adjacency matrices with judicious constraint relaxation. Extensive experimentation demonstrates that FUGAL consistently surpasses state-of-the-art graph alignment methods in accuracy across all benchmark datasets without encumbering efficiency. | FUGAL: Feature-fortified Unrestricted Graph Alignment | [
"Aditya Bommakanti",
"Harshith Reddy Vonteri",
"Konstantinos Skitsas",
"Sayan Ranu",
"Davide Mottin",
"Panagiotis Karras"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SciWuYPNG0 | @inproceedings{
cheng2024information,
title={Information Re-Organization Improves Reasoning in Large Language Models},
author={Xiaoxia Cheng and Zeqi Tan and Wei Xue and Weiming Lu},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SciWuYPNG0}
} | Improving the reasoning capabilities of large language models (LLMs) has attracted considerable interest. Recent approaches primarily focus on improving the reasoning process to yield a more precise final answer. However, in scenarios involving contextually aware reasoning, these methods neglect the importance of first identifying logical relationships from the context before proceeding with the reasoning. This oversight could lead to a superficial understanding and interaction with the context, potentially undermining the quality and reliability of the reasoning outcomes. In this paper, we propose an information re-organization (\textbf{InfoRE}) method before proceeding with the reasoning to enhance the reasoning ability of LLMs. Our re-organization method involves initially extracting logical relationships from the contextual content, such as documents or paragraphs, and subsequently pruning redundant content to minimize noise. Then, we utilize the re-organized information in the reasoning process. This enables LLMs to deeply understand the contextual content by clearly perceiving these logical relationships, while also ensuring high-quality responses by eliminating potential noise. To demonstrate the effectiveness of our approach in improving the reasoning ability, we conduct experiments using Llama2-70B, GPT-3.5, and GPT-4 on various contextually aware multi-hop reasoning tasks. Using only a zero-shot setting, our method achieves an average absolute improvement of 4\% across all tasks, highlighting its potential to improve the reasoning performance of LLMs. | Information Re-Organization Improves Reasoning in Large Language Models | [
"Xiaoxia Cheng",
"Zeqi Tan",
"Wei Xue",
"Weiming Lu"
] | NeurIPS.cc/2024/Conference | 2404.13985 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=ScbmEmtsH5 | @inproceedings{
qu2024discgs,
title={DisC-{GS}: Discontinuity-aware Gaussian Splatting},
author={Haoxuan Qu and Zhuoling Li and Hossein Rahmani and Yujun Cai and Jun Liu},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=ScbmEmtsH5}
} | Recently, Gaussian Splatting, a method that represents a 3D scene as a collection of Gaussian distributions, has gained significant attention in addressing the task of novel view synthesis. In this paper, we highlight a fundamental limitation of Gaussian Splatting: its inability to accurately render discontinuities and boundaries in images due to the continuous nature of Gaussian distributions. To address this issue, we propose a novel framework enabling Gaussian Splatting to perform discontinuity-aware image rendering. Additionally, we introduce a B\'ezier-boundary gradient approximation strategy within our framework to keep the ``differentiability'' of the proposed discontinuity-aware rendering process. Extensive experiments demonstrate the efficacy of our framework. | DisC-GS: Discontinuity-aware Gaussian Splatting | [
"Haoxuan Qu",
"Zhuoling Li",
"Hossein Rahmani",
"Yujun Cai",
"Jun Liu"
] | NeurIPS.cc/2024/Conference | 2405.15196 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SaodQ13jga | @inproceedings{
wei2024gita,
title={{GITA}: Graph to Visual and Textual Integration for Vision-Language Graph Reasoning},
author={Yanbin Wei and Shuai Fu and Weisen Jiang and Zejian Zhang and Zhixiong Zeng and Qi Wu and James Kwok and Yu Zhang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SaodQ13jga}
} | Large Language Models (LLMs) are increasingly used for various tasks with graph structures. Though LLMs can process graph information in a textual format, they overlook the rich vision modality, which is an intuitive way for humans to comprehend structural information and conduct general graph reasoning. The potential benefits and capabilities of representing graph structures as visual images (i.e., $\textit{visual graph}$) are still unexplored. To fill the gap, we innovatively propose an end-to-end framework, called $\textbf{G}$raph to v$\textbf{I}$sual and $\textbf{T}$extual Integr$\textbf{A}$tion (GITA), which firstly incorporates visual graphs into general graph reasoning. Besides, we establish $\textbf{G}$raph-based $\textbf{V}$ision-$\textbf{L}$anguage $\textbf{Q}$uestion $\textbf{A}$nswering (GVLQA) dataset from existing graph data, which is the first vision-language dataset for general graph reasoning purposes. Extensive experiments on the GVLQA dataset and five real-world datasets show that GITA outperforms mainstream LLMs in terms of general graph reasoning capabilities. Moreover, We highlight the effectiveness of the layout augmentation on visual graphs and pretraining on the GVLQA dataset. | GITA: Graph to Visual and Textual Integration for Vision-Language Graph Reasoning | [
"Yanbin Wei",
"Shuai Fu",
"Weisen Jiang",
"Zejian Zhang",
"Zhixiong Zeng",
"Qi Wu",
"James Kwok",
"Yu Zhang"
] | NeurIPS.cc/2024/Conference | 2402.02130 | [
"https://github.com/WEIYanbin1999/GITA"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SadbRPoG2k | @inproceedings{
cohen2024bayesian,
title={Bayesian Strategic Classification},
author={Lee Cohen and Saeed Sharifi -Malvajerdi and Kevin Stangl and Ali Vakilian and Juba Ziani},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SadbRPoG2k}
} | In strategic classification, agents modify their features, at a cost, to obtain a positive classification outcome from the learner’s classifier,
typically assuming agents have full knowledge of the deployed classifier. In contrast, we consider a Bayesian setting where agents have a common distributional prior on the classifier being used and agents manipulate their features to maximize their expected utility according to this prior.
The learner can reveal truthful, yet not necessarily complete, information about the classifier to the agents, aiming to release just enough information to shape the agents' behavior and thus maximize accuracy. We show that partial information release can counter-intuitively benefit the learner’s accuracy, allowing qualified agents to pass the classifier while preventing unqualified agents from doing so. Despite the intractability of computing the best response of an agent in the general case, we provide oracle-efficient algorithms for scenarios where the learner’s hypothesis class consists of low-dimensional linear classifiers or when the agents’ cost function satisfies a sub-modularity condition.
Additionally, we address the learner’s optimization problem, offering both positive and negative results on determining the optimal information release to maximize expected accuracy, particularly in settings where an agent’s qualification can be represented by a real-valued number. | Bayesian Strategic Classification | [
"Lee Cohen",
"Saeed Sharifi -Malvajerdi",
"Kevin Stangl",
"Ali Vakilian",
"Juba Ziani"
] | NeurIPS.cc/2024/Conference | 2402.08758 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SYjxhKcXoN | @inproceedings{
chen2024lfme,
title={{LFME}: A Simple Framework for Learning from Multiple Experts in Domain Generalization},
author={Liang Chen and Yong Zhang and Yibing Song and Zhiqiang Shen and Lingqiao Liu},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SYjxhKcXoN}
} | Domain generalization (DG) methods aim to maintain good performance in an unseen target domain by using training data from multiple source domains. While success on certain occasions are observed, enhancing the baseline across most scenarios remains challenging. This work introduces a simple yet effective framework, dubbed learning from multiple experts (LFME), that aims to make the target model an expert in all source domains to improve DG. Specifically, besides learning the target model used in inference, LFME will also train multiple experts specialized in different domains, whose output probabilities provide professional guidance by simply regularizing the logit of the target model. Delving deep into the framework, we reveal that the introduced logit regularization term implicitly provides effects of enabling the target model to harness more information, and mining hard samples from the experts during training. Extensive experiments on benchmarks from different DG tasks demonstrate that LFME is consistently beneficial to the baseline and can achieve comparable performance to existing arts. Code is available at https://github.com/liangchen527/LFME. | LFME: A Simple Framework for Learning from Multiple Experts in Domain Generalization | [
"Liang Chen",
"Yong Zhang",
"Yibing Song",
"Zhiqiang Shen",
"Lingqiao Liu"
] | NeurIPS.cc/2024/Conference | 2410.17020 | [
"https://github.com/liangchen527/lfme"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SXy1nVGyO7 | @inproceedings{
ye2024on,
title={On the Identifiability of Hybrid Deep Generative Models: Meta-Learning as a Solution},
author={Yubo Ye and Maryam Toloubidokhti and Sumeet Vadhavkar and Xiajun Jiang and Huafeng Liu and Linwei Wang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SXy1nVGyO7}
} | The interest in leveraging physics-based inductive bias in deep learning has resulted in recent development of _hybrid deep generative models (hybrid-DGMs)_ that integrates known physics-based mathematical expressions in neural generative models. To identify these hybrid-DGMs requires inferring parameters of the physics-based component along with their neural component. The identifiability of these hybrid-DGMs, however, has not yet been theoretically probed or established. How does the existing theory of the un-identifiability of general DGMs apply to hybrid-DGMs? What may be an effective approach to consutrct a hybrid-DGM with theoretically-proven identifiability? This paper provides the first theoretical probe into the identifiability of hybrid-DGMs, and present meta-learning as a novel solution to construct identifiable hybrid-DGMs. On synthetic and real-data benchmarks, we provide strong empirical evidence for the un-identifiability of existing hybrid-DGMs using unconditional priors, and strong identifiability results of the presented meta-formulations of hybrid-DGMs. | On the Identifiability of Hybrid Deep Generative Models: Meta-Learning as a Solution | [
"Yubo Ye",
"Maryam Toloubidokhti",
"Sumeet Vadhavkar",
"Xiajun Jiang",
"Huafeng Liu",
"Linwei Wang"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SXbyy0a3rY | @inproceedings{
lee2024groundit,
title={GrounDiT: Grounding Diffusion Transformers via Noisy Patch Transplantation},
author={Yuseung Lee and TaeHoon Yoon and Minhyuk Sung},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SXbyy0a3rY}
} | We introduce GrounDiT, a novel training-free spatial grounding technique for text-to-image generation using Diffusion Transformers (DiT). Spatial grounding with bounding boxes has gained attention for its simplicity and versatility, allowing for enhanced user control in image generation. However, prior training-free approaches often rely on updating the noisy image during the reverse diffusion process via backpropagation from custom loss functions, which frequently struggle to provide precise control over individual bounding boxes. In this work, we leverage the flexibility of the Transformer architecture, demonstrating that DiT can generate noisy patches corresponding to each bounding box, fully encoding the target object and allowing for fine-grained control over each region. Our approach builds on an intriguing property of DiT, which we refer to as semantic sharing. Due to semantic sharing, when a smaller patch is jointly denoised alongside a generatable-size image, the two become semantic clones. Each patch is denoised in its own branch of the generation process and then transplanted into the corresponding region of the original noisy image at each timestep, resulting in robust spatial grounding for each bounding box. In our experiments on the HRS and DrawBench benchmarks, we achieve state-of-the-art performance compared to previous training-free approaches. Project Page: https://groundit-diffusion.github.io/. | GrounDiT: Grounding Diffusion Transformers via Noisy Patch Transplantation | [
"Yuseung Lee",
"TaeHoon Yoon",
"Minhyuk Sung"
] | NeurIPS.cc/2024/Conference | 2410.20474 | [
""
] | https://huggingface.co/papers/2410.20474 | 1 | 13 | 2 | 3 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=STrpbhrvt3 | @inproceedings{
yang2024a,
title={A Textbook Remedy for Domain Shifts: Knowledge Priors for Medical Image Analysis},
author={Yue Yang and Mona Gandhi and Yufei Wang and Yifan Wu and Michael S Yao and Chris Callison-Burch and James Gee and Mark Yatskar},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=STrpbhrvt3}
} | While deep networks have achieved broad success in analyzing natural images, when applied to medical scans, they often fail in unexcepted situations. We investigate this challenge and focus on model sensitivity to domain shifts, such as data sampled from different hospitals or data confounded by demographic variables such as sex, race, etc, in the context of chest X-rays and skin lesion images. A key finding we show empirically is that existing visual backbones lack an appropriate prior from the architecture for reliable generalization in these settings. Taking inspiration from medical training, we propose giving deep networks a prior grounded in explicit medical knowledge communicated in natural language. To this end, we introduce Knowledge-enhanced Bottlenecks (KnoBo), a class of concept bottleneck models that incorporates knowledge priors that constrain it to reason with clinically relevant factors found in medical textbooks or PubMed. KnoBo uses retrieval-augmented language models to design an appropriate concept space paired with an automatic training procedure for recognizing the concept. We evaluate different resources of knowledge and recognition architectures on a broad range of domain shifts across 20 datasets. In our comprehensive evaluation with two imaging modalities, KnoBo outperforms fine-tuned models on confounded datasets by 32.4% on average. Finally, evaluations reveal that PubMed is a promising resource for making medical models less sensitive to domain shift, outperforming other resources on both diversity of information and final prediction performance. | A Textbook Remedy for Domain Shifts: Knowledge Priors for Medical Image Analysis | [
"Yue Yang",
"Mona Gandhi",
"Yufei Wang",
"Yifan Wu",
"Michael S Yao",
"Chris Callison-Burch",
"James Gee",
"Mark Yatskar"
] | NeurIPS.cc/2024/Conference | 2405.14839 | [
""
] | https://huggingface.co/papers/2405.14839 | 1 | 0 | 0 | 8 | [
"yyupenn/whyxrayclip",
"yyupenn/whylesionclip"
] | [] | [] | [
"yyupenn/whyxrayclip",
"yyupenn/whylesionclip"
] | [] | [] | 1 | oral |
null | https://openreview.net/forum?id=SSCtCq2MH2 | @inproceedings{
cai2024gic,
title={{GIC}: Gaussian-Informed Continuum for Physical Property Identification and Simulation},
author={Junhao Cai and Yuji Yang and Weihao Yuan and Yisheng HE and Zilong Dong and Liefeng Bo and Hui Cheng and Qifeng Chen},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SSCtCq2MH2}
} | This paper studies the problem of estimating physical properties (system identification) through visual observations. To facilitate geometry-aware guidance in physical property estimation, we introduce a novel hybrid framework that leverages 3D Gaussian representation to not only capture explicit shapes but also enable the simulated continuum to render object masks as 2D shape surrogates during training. We propose a new dynamic 3D Gaussian framework based on motion factorization to recover the object as 3D Gaussian point sets across different time states. Furthermore, we develop a coarse-to-fine filling strategy to generate the density fields of the object from the Gaussian reconstruction, allowing for the extraction of object continuums along with their surfaces and the integration of Gaussian attributes into these continuum. In addition to the extracted object surfaces, the Gaussian-informed continuum also enables the rendering of object masks during simulations, serving as 2D-shape guidance for physical property estimation. Extensive experimental evaluations demonstrate that our pipeline achieves state-of-the-art performance across multiple benchmarks and metrics. Additionally, we illustrate the effectiveness of the proposed method through real-world demonstrations, showcasing its practical utility. Our project page is at https://jukgei.github.io/project/gic. | GIC: Gaussian-Informed Continuum for Physical Property Identification and Simulation | [
"Junhao Cai",
"Yuji Yang",
"Weihao Yuan",
"Yisheng HE",
"Zilong Dong",
"Liefeng Bo",
"Hui Cheng",
"Qifeng Chen"
] | NeurIPS.cc/2024/Conference | 2406.14927 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=SRWs2wxNs7 | @inproceedings{
tian2024udits,
title={U-DiTs: Downsample Tokens in U-Shaped Diffusion Transformers},
author={Yuchuan Tian and Zhijun Tu and Hanting Chen and Jie Hu and Chao Xu and Yunhe Wang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SRWs2wxNs7}
} | Diffusion Transformers (DiTs) introduce the transformer architecture to diffusion tasks for latent-space image generation. With an isotropic architecture that chains a series of transformer blocks, DiTs demonstrate competitive performance and good scalability; but meanwhile, the abandonment of U-Net by DiTs and their following improvements is worth rethinking. To this end, we conduct a simple toy experiment by comparing a U-Net architectured DiT with an isotropic one. It turns out that the U-Net architecture only gain a slight advantage amid the U-Net inductive bias, indicating potential redundancies within the U-Net-style DiT. Inspired by the discovery that U-Net backbone features are low-frequency-dominated, we perform token downsampling on the query-key-value tuple for self-attention and bring further improvements despite a considerable amount of reduction in computation. Based on self-attention with downsampled tokens, we propose a series of U-shaped DiTs (U-DiTs) in the paper and conduct extensive experiments to demonstrate the extraordinary performance of U-DiT models. The proposed U-DiT could outperform DiT-XL with only 1/6 of its computation cost. Codes are available at https://github.com/YuchuanTian/U-DiT. | U-DiTs: Downsample Tokens in U-Shaped Diffusion Transformers | [
"Yuchuan Tian",
"Zhijun Tu",
"Hanting Chen",
"Jie Hu",
"Chao Xu",
"Yunhe Wang"
] | NeurIPS.cc/2024/Conference | 2405.02730 | [
"https://github.com/yuchuantian/u-dit"
] | https://huggingface.co/papers/2405.02730 | 0 | 0 | 0 | 6 | [
"yuchuantian/U-DiT"
] | [
"yuchuantian/imagenet_vae_256"
] | [] | [
"yuchuantian/U-DiT"
] | [
"yuchuantian/imagenet_vae_256"
] | [] | 1 | poster |
null | https://openreview.net/forum?id=SQicD307Oh | @inproceedings{
chen2024statefree,
title={State-free Reinforcement Learning},
author={Mingyu Chen and Aldo Pacchiano and Xuezhou Zhang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SQicD307Oh}
} | In this work, we study the \textit{state-free RL} problem, where the algorithm does not have the states information before interacting with the environment. Specifically, denote the reachable state set by $\mathcal{S}^\Pi := \{ s|\max_{\pi\in \Pi}q^{P, \pi}(s)>0 \}$, we design an algorithm which requires no information on the state space $S$ while having a regret that is completely independent of $\mathcal{S}$ and only depend on $\mathcal{S}^\Pi$. We view this as a concrete first step towards \textit{parameter-free RL}, with the goal of designing RL algorithms that require no hyper-parameter tuning. | State-free Reinforcement Learning | [
"Mingyu Chen",
"Aldo Pacchiano",
"Xuezhou Zhang"
] | NeurIPS.cc/2024/Conference | 2409.18439 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SQVns9hWJT | @inproceedings{
zeng2024textctrl,
title={TextCtrl: Diffusion-based Scene Text Editing with Prior Guidance Control},
author={Weichao Zeng and Yan Shu and Zhenhang Li and Dongbao Yang and Yu Zhou},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SQVns9hWJT}
} | Centred on content modification and style preservation, Scene Text Editing (STE) remains a challenging task despite considerable progress in text-to-image synthesis and text-driven image manipulation recently. GAN-based STE methods generally encounter a common issue of model generalization, while Diffusion-based STE methods suffer from undesired style deviations. To address these problems, we propose TextCtrl, a diffusion-based method that edits text with prior guidance control. Our method consists of two key components: (i) By constructing fine-grained text style disentanglement and robust text glyph structure representation, TextCtrl explicitly incorporates Style-Structure guidance into model design and network training, significantly improving text style consistency and rendering accuracy. (ii) To further leverage the style prior, a Glyph-adaptive Mutual Self-attention mechanism is proposed which deconstructs the implicit fine-grained features of the source image to enhance style consistency and vision quality during inference. Furthermore, to fill the vacancy of the real-world STE evaluation benchmark, we create the first real-world image-pair dataset termed ScenePair for fair comparisons. Experiments demonstrate the effectiveness of TextCtrl compared with previous methods concerning both style fidelity and text accuracy. Project page: https://github.com/weichaozeng/TextCtrl. | TextCtrl: Diffusion-based Scene Text Editing with Prior Guidance Control | [
"Weichao Zeng",
"Yan Shu",
"Zhenhang Li",
"Dongbao Yang",
"Yu Zhou"
] | NeurIPS.cc/2024/Conference | 2410.10133 | [
"https://github.com/weichaozeng/textctrl"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=SOsiObSdU2 | @inproceedings{
holt2024automatically,
title={Automatically Learning Hybrid Digital Twins of Dynamical Systems},
author={Samuel Holt and Tennison Liu and Mihaela van der Schaar},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SOsiObSdU2}
} | Digital Twins (DTs) are computational models that simulate the states and temporal dynamics of real-world systems, playing a crucial role in prediction, understanding, and decision-making across diverse domains. However, existing approaches to DTs often struggle to generalize to unseen conditions in data-scarce settings, a crucial requirement for such models. To address these limitations, our work begins by establishing the essential desiderata for effective DTs. Hybrid Digital Twins (**HDTwins**) represent a promising approach to address these requirements, modeling systems using a composition of both mechanistic and neural components. This hybrid architecture simultaneously leverages (partial) domain knowledge and neural network expressiveness to enhance generalization, with its modular design facilitating improved evolvability. While existing hybrid models rely on expert-specified architectures with only parameters optimized on data, *automatically* specifying and optimizing HDTwins remains intractable due to the complex search space and the need for flexible integration of domain priors. To overcome this complexity, we propose an evolutionary algorithm (**HDTwinGen**) that employs Large Language Models (LLMs) to autonomously propose, evaluate, and optimize HDTwins. Specifically, LLMs iteratively generate novel model specifications, while offline tools are employed to optimize emitted parameters. Correspondingly, proposed models are evaluated and evolved based on targeted feedback, enabling the discovery of increasingly effective hybrid models. Our empirical results reveal that HDTwinGen produces generalizable, sample-efficient, and evolvable models, significantly advancing DTs' efficacy in real-world applications. | Automatically Learning Hybrid Digital Twins of Dynamical Systems | [
"Samuel Holt",
"Tennison Liu",
"Mihaela van der Schaar"
] | NeurIPS.cc/2024/Conference | 2410.23691 | [
"https://github.com/samholt/HDTwinGen"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | oral |
|
null | https://openreview.net/forum?id=SO7fnIFq0o | @inproceedings{
janz2024ensemble,
title={Ensemble sampling for linear bandits: small ensembles suffice},
author={David Janz and Alexander Litvak and Csaba Szepesvari},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SO7fnIFq0o}
} | We provide the first useful and rigorous analysis of ensemble sampling for the stochastic linear bandit setting. In particular, we show that, under standard assumptions, for a $d$-dimensional stochastic linear bandit with an interaction horizon $T$, ensemble sampling with an ensemble of size of order $\smash{d \log T}$ incurs regret at most of the order $\smash{(d \log T)^{5/2} \sqrt{T}}$. Ours is the first result in any structured setting not to require the size of the ensemble to scale linearly with $T$---which defeats the purpose of ensemble sampling---while obtaining near $\smash{\sqrt{T}}$ order regret. Ours is also the first result that allows infinite action sets. | Ensemble sampling for linear bandits: small ensembles suffice | [
"David Janz",
"Alexander Litvak",
"Csaba Szepesvari"
] | NeurIPS.cc/2024/Conference | 2311.08376 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SO1aRpwVLk | @inproceedings{
yu2024real,
title={4Real: Towards Photorealistic 4D Scene Generation via Video Diffusion Models},
author={Heng Yu and Chaoyang Wang and Peiye Zhuang and Willi Menapace and Aliaksandr Siarohin and Junli Cao and Laszlo Attila Jeni and Sergey Tulyakov and Hsin-Ying Lee},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SO1aRpwVLk}
} | Existing dynamic scene generation methods mostly rely on distilling knowledge from pre-trained 3D generative models, which are typically fine-tuned on synthetic object datasets.
As a result, the generated scenes are often object-centric and lack photorealism.
To address these limitations, we introduce a novel pipeline designed for photorealistic text-to-4D scene generation, discarding the dependency on multi-view generative models and instead fully utilizing video generative models trained on diverse real-world datasets.
Our method begins by generating a reference video using the video generation model.
We then learn the canonical 3D representation of the video using a freeze-time video, delicately generated from the reference video.
To handle inconsistencies in the freeze-time video, we jointly learn a per-frame deformation to model these imperfections.
We then learn the temporal deformation based on the canonical representation to capture dynamic interactions in the reference video.
The pipeline facilitates the generation of dynamic scenes with enhanced photorealism and structural integrity, viewable from multiple perspectives, thereby setting a new standard in 4D scene generation. | 4Real: Towards Photorealistic 4D Scene Generation via Video Diffusion Models | [
"Heng Yu",
"Chaoyang Wang",
"Peiye Zhuang",
"Willi Menapace",
"Aliaksandr Siarohin",
"Junli Cao",
"Laszlo Attila Jeni",
"Sergey Tulyakov",
"Hsin-Ying Lee"
] | NeurIPS.cc/2024/Conference | 2406.07472 | [
""
] | https://huggingface.co/papers/2406.07472 | 7 | 11 | 2 | 9 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=SNmuKbU0am | @inproceedings{
cheng2024mixtures,
title={Mixtures of Experts for Audio-Visual Learning},
author={Ying Cheng and Yang Li and Junjie He and Rui Feng},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SNmuKbU0am}
} | With the rapid development of multimedia technology, audio-visual learning has emerged as a promising research topic within the field of multimodal analysis. In this paper, we explore parameter-efficient transfer learning for audio-visual learning and propose the Audio-Visual Mixture of Experts (\ourmethodname) to inject adapters into pre-trained models flexibly. Specifically, we introduce unimodal and cross-modal adapters as multiple experts to specialize in intra-modal and inter-modal information, respectively, and employ a lightweight router to dynamically allocate the weights of each expert according to the specific demands of each task. Extensive experiments demonstrate that our proposed approach \ourmethodname achieves superior performance across multiple audio-visual tasks,
including AVE, AVVP, AVS, and AVQA. Furthermore, visual-only experimental results also indicate that our approach can tackle challenging scenes where modality information is missing.
The source code is available at \url{https://github.com/yingchengy/AVMOE}. | Mixtures of Experts for Audio-Visual Learning | [
"Ying Cheng",
"Yang Li",
"Junjie He",
"Rui Feng"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SM9IWrHz4e | @inproceedings{
boone2024achieving,
title={Achieving Tractable Minimax Optimal Regret in Average Reward {MDP}s},
author={Victor Boone and Zihan Zhang},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SM9IWrHz4e}
} | In recent years, significant attention has been directed towards learning average-reward Markov Decision Processes (MDPs).
However, existing algorithms either suffer from sub-optimal regret guarantees or computational inefficiencies.
In this paper, we present the first *tractable* algorithm with minimax optimal regret of $\mathrm{O}\left(\sqrt{\mathrm{sp}(h^*) S A T \log(SAT)}\right)$ where $\mathrm{sp}(h^*)$ is the span of the optimal bias function $h^*$, $S\times A$ is the size of the state-action space and $T$ the number of learning steps.
Remarkably, our algorithm does not require prior information on $\mathrm{sp}(h^*)$.
Our algorithm relies on a novel subroutine, **P**rojected **M**itigated **E**xtended **V**alue **I**teration (`PMEVI`), to compute bias-constrained optimal policies efficiently.
This subroutine can be applied to various previous algorithms to obtain improved regret bounds. | Achieving Tractable Minimax Optimal Regret in Average Reward MDPs | [
"Victor Boone",
"Zihan Zhang"
] | NeurIPS.cc/2024/Conference | 2406.01234 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SLuZpdMDFg | @inproceedings{
babiloni2024idtod,
title={{ID}-to-3D: Expressive {ID}-guided 3D Heads via Score Distillation Sampling},
author={Francesca Babiloni and Alexandros Lattas and Jiankang Deng and Stefanos Zafeiriou},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SLuZpdMDFg}
} | We propose ID-to-3D, a method to generate identity- and text-guided 3D human heads with disentangled expressions, starting from even a single casually captured ‘in-the-wild’ image of a subject. The foundation of our approach is anchored in compositionality, alongside the use of task-specific 2D diffusion models as priors for optimization. First, we extend a foundational model with a lightweight expression-aware and ID-aware architecture, and create 2D priors for geometric and texture generation, via fine-tuning only 0.2% of its available training parameters. Then, we jointly leverage a neural parametric representation for the expression of each subject and a multi-stage generation of highly detailed geometry and albedo texture. This combination of strong face identity embeddings and our neural representation enables accurate reconstruction of not only facial features but also accessories and hair, and can be meshed to provide render-ready assets for gaming and telepresence. Our results achieve an unprecedented level of id-consistent and high-quality texture and geometry generation, generalizing to a ‘world’ of unseen 3D identities, without relying on large 3D captured datasets of human assets. | ID-to-3D: Expressive ID-guided 3D Heads via Score Distillation Sampling | [
"Francesca Babiloni",
"Alexandros Lattas",
"Jiankang Deng",
"Stefanos Zafeiriou"
] | NeurIPS.cc/2024/Conference | 2405.16570 | [
""
] | https://huggingface.co/papers/2405.16570 | 0 | 0 | 0 | 4 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=SLnsoaY4u1 | @inproceedings{
xu2024provably,
title={Provably Robust Score-Based Diffusion Posterior Sampling for Plug-and-Play Image Reconstruction},
author={Xingyu Xu and Yuejie Chi},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SLnsoaY4u1}
} | In a great number of tasks in science and engineering, the goal is to infer an unknown image from a small number of noisy measurements collected from a known forward model describing certain sensing or imaging modality. Due to resource constraints, this image reconstruction task is often extremely ill-posed, which necessitates the adoption of expressive prior information to regularize the solution space. Score-based diffusion models, thanks to its impressive empirical success, have emerged as an appealing candidate of an expressive prior in image reconstruction. In order to accommodate diverse tasks at once, it is of great interest to develop efficient, consistent and robust algorithms that incorporate unconditional score functions of an image prior distribution in conjunction with flexible choices of forward models.
This work develops an algorithmic framework for employing score-based diffusion models as an expressive data prior in nonlinear inverse problems with general forward models. Motivated by the plug-and-play framework in the imaging community, we introduce a diffusion plug-and-play method (DPnP) that alternatively calls two samplers, a proximal consistency sampler based solely on the likelihood function of the forward model, and a denoising diffusion sampler based solely on the score functions of the image prior. The key insight is that denoising under white Gaussian noise can be solved rigorously via both stochastic (i.e., DDPM-type) and deterministic (i.e., DDIM-type) samplers using the same set of score functions trained for generation. We establish both asymptotic and non-asymptotic performance guarantees of DPnP, and provide numerical experiments to illustrate its promise in solving both linear and nonlinear image reconstruction tasks. To the best of our knowledge, DPnP is the first provably-robust posterior sampling method for nonlinear inverse problems using unconditional diffusion priors. | Provably Robust Score-Based Diffusion Posterior Sampling for Plug-and-Play Image Reconstruction | [
"Xingyu Xu",
"Yuejie Chi"
] | NeurIPS.cc/2024/Conference | 2403.17042 | [
"https://github.com/x1xu/diffusion-plug-and-play"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SKhR5CuiqQ | @inproceedings{
savani2024diffusing,
title={Diffusing Differentiable Representations},
author={Yash Savani and Marc Anton Finzi and J Zico Kolter},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SKhR5CuiqQ}
} | We introduce a novel, training-free method for sampling *differentiable representations* (diffreps) using pretrained diffusion models. Rather than merely mode-seeking, our method achieves sampling by "pulling back" the dynamics of the reverse-time process—from the image space to the diffrep parameter space—and updating the parameters according to this pulled-back process. We identify an implicit constraint on the samples induced by the diffrep and demonstrate that addressing this constraint significantly improves the consistency and detail of the generated objects. Our method yields diffreps with substantially **improved quality and diversity** for images, panoramas, and 3D NeRFs compared to existing techniques. Our approach is a general-purpose method for sampling diffreps, expanding the scope of problems that diffusion models can tackle. | Diffusing Differentiable Representations | [
"Yash Savani",
"Marc Anton Finzi",
"J Zico Kolter"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SKY1ScUTwA | @inproceedings{
bechler-speicher2024the,
title={The Intelligible and Effective Graph Neural Additive Network},
author={Maya Bechler-Speicher and Amir Globerson and Ran Gilad-Bachrach},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SKY1ScUTwA}
} | Graph Neural Networks (GNNs) have emerged as the predominant approach for learning over graph-structured data. However, most GNNs operate as black-box models and require post-hoc explanations, which may not suffice in high-stakes scenarios where transparency is crucial.
In this paper, we present a GNN that is interpretable by design. Our model, Graph Neural Additive Network (GNAN), is a novel extension of the interpretable class of Generalized Additive Models, and can be visualized and fully understood by humans. GNAN is designed to be fully interpretable, offering both global and local explanations at the feature and graph levels through direct visualization of the model. These visualizations describe exactly how the model uses the relationships between the target variable, the features, and the graph. We demonstrate the intelligibility of GNANs in a series of examples on different tasks and datasets. In addition, we show that the accuracy of GNAN is on par with black-box GNNs, making it suitable for critical applications where transparency is essential, alongside high accuracy. | The Intelligible and Effective Graph Neural Additive Network | [
"Maya Bechler-Speicher",
"Amir Globerson",
"Ran Gilad-Bachrach"
] | NeurIPS.cc/2024/Conference | [
"https://github.com/mayabechlerspeicher/Graph-Neural-Additive-Networks---GNAN"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SKCbZR8Pyd | @inproceedings{
zhang2024speechalign,
title={SpeechAlign: Aligning Speech Generation to Human Preferences},
author={Dong Zhang and Zhaowei Li and Shimin Li and Xin Zhang and Pengyu Wang and Yaqian Zhou and Xipeng Qiu},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SKCbZR8Pyd}
} | Speech language models have significantly advanced in generating realistic speech, with neural codec language models standing out. However, the integration of preference optimization to align speech outputs to human preferences is often neglected. This paper addresses this gap by first analyzing the distribution gap in codec language models, highlighting how it leads to discrepancies between the training and inference phases, which negatively affects performance. Then we explore leveraging preference optimization to bridge the distribution gap. We introduce SpeechAlign, an iterative self-improvement strategy that aligns speech language models to human preferences. SpeechAlign involves constructing a preference codec dataset contrasting golden codec tokens against synthetic tokens, followed by preference optimization to improve the codec language model. This cycle of improvement is carried out iteratively to steadily convert weak models to strong ones. Through both subjective and objective evaluations, we show that SpeechAlign can bridge the distribution gap and facilitating continuous self-improvement of the speech language model. Moreover, SpeechAlign exhibits robust generalization capabilities and works for smaller models. Demos are available at https://0nutation.github.io/SpeechAlign.github.io/. | SpeechAlign: Aligning Speech Generation to Human Preferences | [
"Dong Zhang",
"Zhaowei Li",
"Shimin Li",
"Xin Zhang",
"Pengyu Wang",
"Yaqian Zhou",
"Xipeng Qiu"
] | NeurIPS.cc/2024/Conference | 2404.05600 | [
"https://github.com/0nutation/speechgpt"
] | https://huggingface.co/papers/2404.05600 | 0 | 1 | 0 | 7 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=SGcnphYOeq | @inproceedings{
takezawa2024parameterfree,
title={Parameter-free Clipped Gradient Descent Meets Polyak},
author={Yuki Takezawa and Han Bao and Ryoma Sato and Kenta Niwa and Makoto Yamada},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SGcnphYOeq}
} | Gradient descent and its variants are de facto standard algorithms for training machine learning models. As gradient descent is sensitive to its hyperparameters, we need to tune the hyperparameters carefully using a grid search. However, the method is time-consuming, particularly when multiple hyperparameters exist. Therefore, recent studies have analyzed parameter-free methods that adjust the hyperparameters on the fly. However, the existing work is limited to investigations of parameter-free methods for the stepsize, and parameter-free methods for other hyperparameters have not been explored. For instance, although the gradient clipping threshold is a crucial hyperparameter in addition to the stepsize for preventing gradient explosion issues, none of the existing studies have investigated parameter-free methods for clipped gradient descent. Therefore, in this study, we investigate the parameter-free methods for clipped gradient descent. Specifically, we propose Inexact Polyak Stepsize, which converges to the optimal solution without any hyperparameters tuning, and its convergence rate is asymptotically independent of $L$ under $L$-smooth and $(L_0, L_1)$-smooth assumptions of the loss function, similar to that of clipped gradient descent with well-tuned hyperparameters. We numerically validated our convergence results using a synthetic function and demonstrated the effectiveness of our proposed methods using LSTM, Nano-GPT, and T5. | Parameter-free Clipped Gradient Descent Meets Polyak | [
"Yuki Takezawa",
"Han Bao",
"Ryoma Sato",
"Kenta Niwa",
"Makoto Yamada"
] | NeurIPS.cc/2024/Conference | 2405.15010 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SFxAjB7UXx | @inproceedings{
large2024scalable,
title={Scalable Optimization in the Modular Norm},
author={Tim Large and Yang Liu and Minyoung Huh and Hyojin Bahng and Phillip Isola and Jeremy Bernstein},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SFxAjB7UXx}
} | To improve performance in contemporary deep learning, one is interested in scaling up the neural network in terms of both the number and the size of the layers. When ramping up the width of a single layer, graceful scaling of training has been linked to the need to normalize the weights and their updates in the "natural norm" particular to that layer. In this paper, we significantly generalize this idea by defining the modular norm, which is the natural norm on the full weight space of any neural network architecture. The modular norm is defined recursively in tandem with the network architecture itself. We show that the modular norm has several promising applications. On the practical side, the modular norm can be used to normalize the updates of any base optimizer so that the learning rate becomes transferable across width and depth. This means that the user does not need to compute optimizer-specific scale factors in order to scale training. On the theoretical side, we show that for any neural network built from "well-behaved" atomic modules, the gradient of the network is Lipschitz-continuous in the modular norm, with the Lipschitz constant admitting a simple recursive formula. This characterization opens the door to porting standard ideas in optimization theory over to deep learning. We have created a Python package called Modula that automatically normalizes weight updates in the modular norm of the architecture. Both the Modula package and code for our experiments are provided in the supplementary material. | Scalable Optimization in the Modular Norm | [
"Tim Large",
"Yang Liu",
"Minyoung Huh",
"Hyojin Bahng",
"Phillip Isola",
"Jeremy Bernstein"
] | NeurIPS.cc/2024/Conference | 2405.14813 | [
"https://github.com/jxbz/modula"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SFk7AMpyhx | @inproceedings{
zhang2024diffusion,
title={4Diffusion: Multi-view Video Diffusion Model for 4D Generation},
author={Haiyu Zhang and Xinyuan Chen and Yaohui Wang and Xihui Liu and Yunhong Wang and Yu Qiao},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SFk7AMpyhx}
} | Current 4D generation methods have achieved noteworthy efficacy with the aid of advanced diffusion generative models. However, these methods lack multi-view spatial-temporal modeling and encounter challenges in integrating diverse prior knowledge from multiple diffusion models, resulting in inconsistent temporal appearance and flickers. In this paper, we propose a novel 4D generation pipeline, namely $\textbf{4Diffusion}$, aimed at generating spatial-temporally consistent 4D content from a monocular video. We first design a unified diffusion model tailored for multi-view video generation by incorporating a learnable motion module into a frozen 3D-aware diffusion model to capture multi-view spatial-temporal correlations. After training on a curated dataset, our diffusion model acquires reasonable temporal consistency and inherently preserves the generalizability and spatial consistency of the 3D-aware diffusion model. Subsequently, we propose 4D-aware Score Distillation Sampling loss, which is based on our multi-view video diffusion model, to optimize 4D representation parameterized by dynamic NeRF. This aims to eliminate discrepancies arising from multiple diffusion models, allowing for generating spatial-temporally consistent 4D content. Moreover, we devise an anchor loss to enhance the appearance details and facilitate the learning of dynamic NeRF. Extensive qualitative and quantitative experiments demonstrate that our method achieves superior performance compared to previous methods. | 4Diffusion: Multi-view Video Diffusion Model for 4D Generation | [
"Haiyu Zhang",
"Xinyuan Chen",
"Yaohui Wang",
"Xihui Liu",
"Yunhong Wang",
"Yu Qiao"
] | NeurIPS.cc/2024/Conference | 2405.20674 | [
""
] | https://huggingface.co/papers/2405.20674 | 4 | 12 | 1 | 6 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=SFaEENfEyw | @inproceedings{
li2024the,
title={The Closeness of In-Context Learning and Weight Shifting for Softmax Regression},
author={Shuai Li and Zhao Song and Yu Xia and Tong Yu and Tianyi Zhou},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SFaEENfEyw}
} | Large language models (LLMs) are known for their exceptional performance in natural language processing, making them highly effective in many human life-related tasks. The attention mechanism in the Transformer architecture is a critical component of LLMs, as it allows the model to selectively focus on specific input parts. The softmax unit, which is a key part of the attention mechanism, normalizes the attention scores. Hence, the performance of LLMs in various NLP tasks depends significantly on the crucial role played by the attention mechanism with the softmax unit.
In-context learning is one of the celebrated abilities of recent LLMs.
Without further parameter updates, Transformers can learn to predict based on few in-context examples.
However, the reason why Transformers becomes in-context learners is not well understood.
Recently, in-context learning has been studied from a mathematical perspective with simplified linear self-attention without softmax unit.
Based on a linear regression formulation $\min_x\| Ax - b \|_2$, existing works show linear Transformers' capability of learning linear functions in context. The capability of Transformers with softmax unit approaching full Transformers, however, remains unexplored.
In this work, we study the in-context learning based on a softmax regression formulation $\min_{x} \| \langle \exp(Ax), {\bf 1}_n \rangle^{-1} \exp(Ax) - b \|_2$. We show the upper bounds of the data transformations induced by a single self-attention layer with softmax unit and by gradient-descent on a $\ell_2$ regression loss for softmax prediction function.
Our theoretical results imply that when training self-attention-only Transformers for fundamental regression tasks, the models learned by gradient-descent and Transformers show great similarity. | The Closeness of In-Context Learning and Weight Shifting for Softmax Regression | [
"Shuai Li",
"Zhao Song",
"Yu Xia",
"Tong Yu",
"Tianyi Zhou"
] | NeurIPS.cc/2024/Conference | 2304.13276 | [
""
] | https://huggingface.co/papers/2304.13276 | 0 | 1 | 0 | 5 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=SFCZdXDyNs | @inproceedings{
jayaraman2024dj,
title={D\'ej\`a Vu Memorization in Vision{\textendash}Language Models},
author={Bargav Jayaraman and Chuan Guo and Kamalika Chaudhuri},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SFCZdXDyNs}
} | Vision-Language Models (VLMs) have emerged as the state-of-the-art representation learning solution, with myriads of downstream applications such as image classification, retrieval and generation. A natural question is whether these models memorize their training data, which also has implications for generalization. We propose a new method for measuring memorization in VLMs, which we call dèjá vu memorization. For VLMs trained on image-caption pairs, we show that the model indeed retains information about individual objects in the training images beyond what can be inferred from correlations or the image caption. We evaluate dèjá vu memorization at both sample and population level, and show that it is significant for OpenCLIP trained on as many as 50M image-caption pairs. Finally, we show that text randomization considerably mitigates memorization risk while only moderately impacting the model’s downstream task performance. The code is available here: https://github.com/facebookresearch/VLMDejaVu. | Déjà Vu Memorization in Vision–Language Models | [
"Bargav Jayaraman",
"Chuan Guo",
"Kamalika Chaudhuri"
] | NeurIPS.cc/2024/Conference | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
||
null | https://openreview.net/forum?id=SF2GlFhVsS | @inproceedings{
chen2024alleviating,
title={Alleviating Hallucinations in Large Vision-Language Models through Hallucination-Induced Optimization},
author={Beitao Chen and Xinyu Lyu and Lianli Gao and Heng Tao Shen and Jingkuan Song},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SF2GlFhVsS}
} | Although Large Visual Language Models (LVLMs) have demonstrated exceptional abilities in understanding multimodal data, they invariably suffer from hallucinations, leading to a disconnection between the generated text and the corresponding images. Almost all current visual contrastive decoding methods attempt to mitigate these hallucinations by introducing visual uncertainty information that appropriately widens the contrastive logits gap between hallucinatory and targeted ones.
However, due to uncontrollable nature of the global visual uncertainty, they struggle to precisely induce the hallucinatory tokens, which severely limits their effectiveness in mitigating hallucinations and may even lead to the generation of undesired hallucinations.
To tackle this issue, we conducted the theoretical analysis to promote the effectiveness of contrast decoding. Building on this insight, we introduce a novel optimization strategy named Hallucination-Induced Optimization (HIO). This strategy seeks to amplify the contrast between hallucinatory and targeted tokens relying on a fine-tuned theoretical preference model (i.e., Contrary Bradley-Terry Model), thereby facilitating efficient contrast decoding to alleviate hallucinations in LVLMs.
Extensive experimental research demonstrates that our HIO strategy can effectively reduce hallucinations in LVLMs, outperforming state-of-the-art methods across various benchmarks. | Alleviating Hallucinations in Large Vision-Language Models through Hallucination-Induced Optimization | [
"Beitao Chen",
"Xinyu Lyu",
"Lianli Gao",
"Heng Tao Shen",
"Jingkuan Song"
] | NeurIPS.cc/2024/Conference | 2405.15356 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SEflLHIhhJ | @inproceedings{
roulet2024stepping,
title={Stepping on the Edge: Curvature Aware Learning Rate Tuners},
author={Vincent Roulet and Atish Agarwala and Jean-Bastien Grill and Grzegorz Michal Swirszcz and Mathieu Blondel and Fabian Pedregosa},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SEflLHIhhJ}
} | Curvature information -- particularly, the largest eigenvalue of the loss
Hessian, known as the sharpness -- often forms the basis for learning rate
tuners. However, recent work has shown that the curvature information undergoes
complex dynamics during training, going from a phase of increasing sharpness to
eventual stabilization. We analyze the closed-loop feedback effect between
learning rate tuning and curvature. We find that classical learning rate tuners
may yield greater one-step loss reduction, yet they ultimately underperform in
the long term when compared to constant learning rates in the full batch regime.
These models break the stabilization of the sharpness, which we explain using a
simplified model of the joint dynamics of the learning rate and the curvature.
To further investigate these effects, we introduce a new learning rate tuning
method, Curvature Dynamics Aware Tuning (CDAT), which prioritizes long term
curvature stabilization over instantaneous progress on the objective. In the
full batch regime, CDAT shows behavior akin to prefixed warm-up schedules on deep
learning objectives, outperforming tuned constant learning rates. In the mini
batch regime, we observe that stochasticity introduces confounding effects that
explain the previous success of some learning rate tuners at appropriate batch
sizes. Our findings highlight the critical role of understanding the joint
dynamics of the learning rate and curvature, beyond greedy minimization, to
diagnose failures and design effective adaptive learning rate tuners. | Stepping on the Edge: Curvature Aware Learning Rate Tuners | [
"Vincent Roulet",
"Atish Agarwala",
"Jean-Bastien Grill",
"Grzegorz Michal Swirszcz",
"Mathieu Blondel",
"Fabian Pedregosa"
] | NeurIPS.cc/2024/Conference | 2407.06183 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SDWeIGPAh9 | @inproceedings{
liu2024typicalnessaware,
title={Typicalness-Aware Learning for Failure Detection},
author={Yijun Liu and Jiequan Cui and Zhuotao Tian and Senqiao Yang and Qingdong He and wangxiaoling and Jingyong Su},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SDWeIGPAh9}
} | Deep neural networks (DNNs) often suffer from the overconfidence issue, where incorrect predictions are made with high confidence scores, hindering the applications in critical systems. In this paper, we propose a novel approach called Typicalness-Aware Learning (TAL) to address this issue and improve failure detection performance.
We observe that, with the cross-entropy loss, model predictions are optimized to align with the corresponding labels via increasing logit magnitude or refining logit direction. However, regarding atypical samples, the image content and their labels may exhibit disparities. This discrepancy can lead to overfitting on atypical samples, ultimately resulting in the overconfidence issue that we aim to address.
To address this issue, we have devised a metric that quantifies the typicalness of each sample, enabling the dynamic adjustment of the logit magnitude during the training process. By allowing relatively atypical samples to be adequately fitted while preserving reliable logit direction, the problem of overconfidence can be mitigated. TAL has been extensively evaluated on benchmark datasets, and the results demonstrate its superiority over existing failure detection methods. Specifically, TAL achieves a more than 5\% improvement on CIFAR100 in terms of the Area Under the Risk-Coverage Curve (AURC) compared to the state-of-the-art. Code is available at https://github.com/liuyijungoon/TAL. | Typicalness-Aware Learning for Failure Detection | [
"Yijun Liu",
"Jiequan Cui",
"Zhuotao Tian",
"Senqiao Yang",
"Qingdong He",
"wangxiaoling",
"Jingyong Su"
] | NeurIPS.cc/2024/Conference | 2411.01981 | [
"https://github.com/liuyijungoon/tal"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SCEdoGghcw | @inproceedings{
karvonen2024measuring,
title={Measuring Progress in Dictionary Learning for Language Model Interpretability with Board Game Models},
author={Adam Karvonen and Benjamin Wright and Can Rager and Rico Angell and Jannik Brinkmann and Logan Riggs Smith and Claudio Mayrink Verdun and David Bau and Samuel Marks},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SCEdoGghcw}
} | What latent features are encoded in language model (LM) representations? Recent work on training sparse autoencoders (SAEs) to disentangle interpretable features in LM representations has shown significant promise. However, evaluating the quality of these SAEs is difficult because we lack a ground-truth collection of interpretable features which we expect good SAEs to identify. We thus propose to measure progress in interpretable dictionary learning by working in the setting of LMs trained on Chess and Othello transcripts. These settings carry natural collections of interpretable features—for example, “there is a knight on F3”—which we leverage into metrics for SAE quality. To guide progress in interpretable dictionary learning, we introduce a new SAE training technique, $p$-annealing, which demonstrates improved performance on our metric. | Measuring Progress in Dictionary Learning for Language Model Interpretability with Board Game Models | [
"Adam Karvonen",
"Benjamin Wright",
"Can Rager",
"Rico Angell",
"Jannik Brinkmann",
"Logan Riggs Smith",
"Claudio Mayrink Verdun",
"David Bau",
"Samuel Marks"
] | NeurIPS.cc/2024/Conference | 2408.00113 | [
"https://github.com/adamkarvonen/SAE_BoardGameEval"
] | https://huggingface.co/papers/2408.00113 | 4 | 6 | 2 | 9 | [] | [] | [] | [] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=SAZeQV2PtT | @inproceedings{
campbell2024general,
title={General bounds on the quality of Bayesian coresets},
author={Trevor Campbell},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SAZeQV2PtT}
} | Bayesian coresets speed up posterior inference in the large-scale data regime by approximating the full-data log-likelihood function with a surrogate log-likelihood based on a small, weighted subset of the data. But while Bayesian coresets and methods for construction are applicable in a wide range of models, existing theoretical analysis of the posterior inferential error incurred by coreset approximations only apply in restrictive settings---i.e., exponential family models, or models with strong log-concavity and smoothness assumptions. This work presents general upper and lower bounds on the Kullback-Leibler (KL) divergence of coreset approximations that reflect the full range of applicability of Bayesian coresets. The lower bounds require only mild model assumptions typical of Bayesian asymptotic analyses, while the upper bounds require the log-likelihood functions to satisfy a generalized subexponentiality criterion that is weaker than conditions used in earlier work. The lower bounds are applied to obtain fundamental limitations on the quality of coreset approximations, and to provide a theoretical explanation for the previously-observed poor empirical performance of importance sampling-based construction methods. The upper bounds are used to analyze the performance of recent subsample-optimize methods. The flexibility of the theory is demonstrated in validation experiments involving multimodal, unidentifiable, heavy-tailed Bayesian posterior distributions. | General bounds on the quality of Bayesian coresets | [
"Trevor Campbell"
] | NeurIPS.cc/2024/Conference | 2405.11780 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=SAQXbnvv4t | @inproceedings{
song2024alchemistcoder,
title={AlchemistCoder: Harmonizing and Eliciting Code Capability by Hindsight Tuning on Multi-source Data},
author={Zifan Song and Yudong Wang and Wenwei Zhang and Kuikun Liu and Chengqi Lyu and Demin Song and Qipeng Guo and Hang Yan and Dahua Lin and Kai Chen and Cairong Zhao},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=SAQXbnvv4t}
} | Open-source Large Language Models (LLMs) and their specialized variants, particularly Code LLMs, have recently delivered impressive performance. However, previous Code LLMs are typically fine-tuned on single-source data with limited quality and diversity, which may insufficiently elicit the potential of pre-trained Code LLMs. In this paper, we present AlchemistCoder, a series of Code LLMs with enhanced code generation and generalization capabilities fine-tuned on multi-source data. To achieve this, we pioneer to unveil inherent conflicts among the various styles and qualities in multi-source code corpora and introduce data-specific prompts with hindsight relabeling, termed AlchemistPrompts, to harmonize different data sources and instruction-response pairs. Additionally, we propose incorporating the data construction process into the fine-tuning data as code comprehension tasks, including instruction evolution, data filtering, and code review. Extensive experiments demonstrate that AlchemistCoder holds a clear lead among all models of the same size (6.7B/7B) and rivals or even surpasses larger models (15B/33B/70B), showcasing the efficacy of our method in refining instruction-following capabilities and advancing the boundaries of code intelligence. Source code and models are available at https://github.com/InternLM/AlchemistCoder. | AlchemistCoder: Harmonizing and Eliciting Code Capability by Hindsight Tuning on Multi-source Data | [
"Zifan Song",
"Yudong Wang",
"Wenwei Zhang",
"Kuikun Liu",
"Chengqi Lyu",
"Demin Song",
"Qipeng Guo",
"Hang Yan",
"Dahua Lin",
"Kai Chen",
"Cairong Zhao"
] | NeurIPS.cc/2024/Conference | 2405.19265 | [
"https://github.com/internlm/alchemistcoder"
] | https://huggingface.co/papers/2405.19265 | 1 | 0 | 0 | 11 | [
"internlm/AlchemistCoder-DS-6.7B",
"lmstudio-community/AlchemistCoder-L-7B-GGUF",
"internlm/AlchemistCoder-L-7B",
"lmstudio-community/AlchemistCoder-DS-6.7B-GGUF",
"internlm/AlchemistCoder-CL-7B",
"QuantFactory/AlchemistCoder-L-7B-GGUF",
"cgus/AlchemistCoder-DS-6.7B-exl2",
"RichardErkhov/internlm_-_AlchemistCoder-L-7B-gguf"
] | [] | [] | [
"internlm/AlchemistCoder-DS-6.7B",
"lmstudio-community/AlchemistCoder-L-7B-GGUF",
"internlm/AlchemistCoder-L-7B",
"lmstudio-community/AlchemistCoder-DS-6.7B-GGUF",
"internlm/AlchemistCoder-CL-7B",
"QuantFactory/AlchemistCoder-L-7B-GGUF",
"cgus/AlchemistCoder-DS-6.7B-exl2",
"RichardErkhov/internlm_-_AlchemistCoder-L-7B-gguf"
] | [] | [] | 1 | poster |
null | https://openreview.net/forum?id=S98OzJD3jn | @inproceedings{
zhong2024diffusion,
title={Diffusion Tuning: Transferring Diffusion Models via Chain of Forgetting},
author={Jincheng Zhong and Xingzhuo Guo and Jiaxiang Dong and Mingsheng Long},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=S98OzJD3jn}
} | Diffusion models have significantly advanced the field of generative modeling. However, training a diffusion model is computationally expensive, creating a pressing need to adapt off-the-shelf diffusion models for downstream generation tasks. Current fine-tuning methods focus on parameter-efficient transfer learning but overlook the fundamental transfer characteristics of diffusion models.
In this paper, we investigate the transferability of diffusion models and observe a monotonous chain of forgetting trend of transferability along the reverse process. Based on this observation and novel theoretical insights, we present Diff-Tuning, a frustratingly simple transfer approach that leverages the chain of forgetting tendency. Diff-Tuning encourages the fine-tuned model to retain the pre-trained knowledge at the end of the denoising chain close to the generated data while discarding the other noise side.
We conduct comprehensive experiments to evaluate Diff-Tuning, including the transfer of pre-trained Diffusion Transformer models to eight downstream generations and the adaptation of Stable Diffusion to five control conditions with ControlNet.
Diff-Tuning achieves a 24.6% improvement over standard fine-tuning and enhances the convergence speed of ControlNet by 24%. Notably, parameter-efficient transfer learning techniques for diffusion models can also benefit from Diff-Tuning. Code
is available at this repository: https://github.com/thuml/Diffusion-Tuning. | Diffusion Tuning: Transferring Diffusion Models via Chain of Forgetting | [
"Jincheng Zhong",
"Xingzhuo Guo",
"Jiaxiang Dong",
"Mingsheng Long"
] | NeurIPS.cc/2024/Conference | 2406.00773 | [
""
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
|
null | https://openreview.net/forum?id=S93hrwT8u9 | @inproceedings{
nguyen2024activation,
title={Activation Map Compression through Tensor Decomposition for Deep Learning},
author={Le-Trung Nguyen and A{\"e}l Qu{\'e}lennec and Enzo Tartaglione and Samuel Tardieu and Van-Tam Nguyen},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=S93hrwT8u9}
} | Internet of Things and Deep Learning are synergetically and exponentially growing industrial fields with a massive call for their unification into a common framework called Edge AI. While on-device inference is a well-explored topic in recent research, backpropagation remains an open challenge due to its prohibitive computational and memory costs compared to the extreme resource constraints of embedded devices. Drawing on tensor decomposition research, we tackle the main bottleneck of backpropagation, namely the memory footprint of activation map storage. We investigate and compare the effects of activation compression using Singular Value Decomposition and its tensor variant, High-Order Singular Value Decomposition. The application of low-order decomposition results in considerable memory savings while preserving the features essential for learning, and also offers theoretical guarantees to convergence. Experimental results obtained on main-stream architectures and tasks demonstrate Pareto-superiority over other state-of-the-art solutions, in terms of the trade-off between generalization and memory footprint. | Activation Map Compression through Tensor Decomposition for Deep Learning | [
"Le-Trung Nguyen",
"Aël Quélennec",
"Enzo Tartaglione",
"Samuel Tardieu",
"Van-Tam Nguyen"
] | NeurIPS.cc/2024/Conference | 2411.06346 | [
"https://github.com/le-trungnguyen/neurips2024-activationcompression"
] | -1 | -1 | -1 | -1 | [] | [] | [] | [] | [] | [] | 0 | poster |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.