NeurIPSConference/NeurIPS2024-papers · Datasets at Hugging Face

null

https://openreview.net/forum?id=jzngdJQ2lY

@inproceedings{ so2024solving, title={Solving Minimum-Cost Reach Avoid using Reinforcement Learning}, author={Oswin So and Cheng Ge and Chuchu Fan}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jzngdJQ2lY} }

Current reinforcement-learning methods are unable to directly learn policies that solve the minimum cost reach-avoid problem to minimize cumulative costs subject to the constraints of reaching the goal and avoiding unsafe states, as the structure of this new optimization problem is incompatible with current methods. Instead, a surrogate problem is solved where all objectives are combined with a weighted sum. However, this surrogate objective results in suboptimal policies that do not directly minimize the cumulative cost. In this work, we propose RC-PPO, a reinforcement-learning-based method for solving the minimum-cost reach-avoid problem by using connections to Hamilton-Jacobi reachability. Empirical results demonstrate that RC-PPO learns policies with comparable goal-reaching rates to while achieving up to 57% lower cumulative costs compared to existing methods on a suite of minimum-cost reach-avoid benchmarks on the Mujoco simulator. The project page can be found at https://oswinso.xyz/rcppo.

Solving Minimum-Cost Reach Avoid using Reinforcement Learning

[ "Oswin So", "Cheng Ge", "Chuchu Fan" ]

NeurIPS.cc/2024/Conference

2410.22600

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jzkpwcj200

@inproceedings{ polo2024efficient, title={Efficient multi-prompt evaluation of {LLM}s}, author={Felipe Maia Polo and Ronald Xu and Lucas Weber and M{\'\i}rian Silva and Onkar Bhardwaj and Leshem Choshen and Allysson Flavio Melo de Oliveira and Yuekai Sun and Mikhail Yurochkin}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jzkpwcj200} }

Most popular benchmarks for comparing LLMs rely on a limited set of prompt templates, which may not fully capture the LLMs’ abilities and can affect the reproducibility of results on leaderboards. Many recent works empirically verify prompt sensitivity and advocate for changes in LLM evaluation. In this paper, we consider the problem of estimating the performance distribution across many prompt variants instead of finding a single prompt to evaluate with. We introduce PromptEval, a method for estimating performance across a large set of prompts borrowing strength across prompts and examples to produce accurate estimates under practical evaluation budgets. The resulting distribution can be used to obtain performance quantiles to construct various robust performance metrics (e.g., top 95% quantile or median). We prove that PromptEval consistently estimates the performance distribution and demonstrate its efficacy empirically on three prominent LLM benchmarks: MMLU, BIG-bench Hard, and LMentry; for example, PromptEval can accurately estimate performance quantiles across 100 prompt templates on MMLU with a budget equivalent to two single-prompt evaluations. Moreover, we show how PromptEval can be useful in LLM-as-a-judge and best prompt identification applications.

Efficient multi-prompt evaluation of LLMs

[ "Felipe Maia Polo", "Ronald Xu", "Lucas Weber", "Mírian Silva", "Onkar Bhardwaj", "Leshem Choshen", "Allysson Flavio Melo de Oliveira", "Yuekai Sun", "Mikhail Yurochkin" ]

NeurIPS.cc/2024/Conference

2405.17202

[ "https://github.com/microsoft/promptbench" ]

https://huggingface.co/papers/2405.17202

3

2

0

9

[]

[ "PromptEval/PromptEval_MMLU_full", "PromptEval/PromptEval_MMLU_correctness" ]

[]

[ "PromptEval/PromptEval_MMLU_full", "PromptEval/PromptEval_MMLU_correctness" ]

[]

1

poster

null

https://openreview.net/forum?id=jz5ZMeN9He

@inproceedings{ li2024drip, title={{DRIP}: Unleashing Diffusion Priors for Joint Foreground and Alpha Prediction in Image Matting}, author={Xiaodi Li and Zongxin Yang and Ruijie Quan and Yi Yang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jz5ZMeN9He} }

Recovering the foreground color and opacity/alpha matte from a single image (i.e., image matting) is a challenging and ill-posed problem where data priors play a critical role in achieving precise results. Traditional methods generally predict the alpha matte and then extract the foreground through post-processing, often failing to produce high-fidelity foreground color. This failure stems from the models' difficulty in learning robust color predictions from limited matting datasets. To address this, we explore the potential of leveraging vision priors embedded in pre-trained latent diffusion models (LDM) for estimating foreground RGBA values in challenging scenarios and rare objects. We introduce Drip, a novel approach for image matting that harnesses the rich prior knowledge of LDM models. Our method incorporates a switcher and a cross-domain attention mechanism to extend the original LDM for joint prediction of the foreground color and opacity. This setup facilitates mutual information exchange and ensures high consistency across both modalities. To mitigate the inherent reconstruction errors of the LDM's VAE decoder, we propose a latent transparency decoder to align the RGBA prediction with the input image, thereby reducing discrepancies. Comprehensive experimental results demonstrate that our approach achieves state-of-the-art performance in foreground and alpha predictions and shows remarkable generalizability across various benchmarks.

DRIP: Unleashing Diffusion Priors for Joint Foreground and Alpha Prediction in Image Matting

[ "Xiaodi Li", "Zongxin Yang", "Ruijie Quan", "Yi Yang" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jwh9MHEfmY

@inproceedings{ yang2024regularizing, title={Regularizing Hidden States Enables Learning Generalizable Reward Model for {LLM}s}, author={Rui Yang and Ruomeng Ding and Yong Lin and Huan Zhang and Tong Zhang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jwh9MHEfmY} }

Reward models trained on human preference data have been proven to effectively align Large Language Models (LLMs) with human intent within the framework of reinforcement learning from human feedback (RLHF). However, current reward models have limited generalization capabilities to unseen prompts and responses, which can lead to an unexpected phenomenon known as reward over-optimization, resulting in a decline in actual performance due to excessive optimization of rewards. While previous research has advocated for constraining policy optimization, our study introduces a novel approach to enhance the reward model's generalization ability against distribution shifts by regularizing the hidden states. Specifically, we retain the base model's language model head and incorporate a suite of text-generation losses to preserve the hidden states' text-generation capabilities, while concurrently learning a reward head behind the same hidden states. Our experimental results demonstrate that the introduced regularization technique markedly improves the accuracy of learned reward models across a variety of out-of-distribution (OOD) tasks and effectively alleviates the over-optimization issue in RLHF, offering a more reliable and robust preference learning paradigm.

Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs

[ "Rui Yang", "Ruomeng Ding", "Yong Lin", "Huan Zhang", "Tong Zhang" ]

NeurIPS.cc/2024/Conference

2406.10216

[ "https://github.com/yangrui2015/generalizable-reward-model" ]

https://huggingface.co/papers/2406.10216

1

2

0

5

[ "Ray2333/reward-model-Mistral-7B-instruct-Unified-Feedback", "Ray2333/GRM-llama3-8B-distill", "Ray2333/GRM-llama3-8B-sftreg", "Ray2333/GRM-Gemma-2B-sftreg", "Ray2333/GRM-gemma2-2B-rewardmodel-ft", "Ray2333/GRM-Llama3.2-3B-rewardmodel-ft", "Ray2333/GRM-Gemma-2B-rewardmodel-ft", "Ray2333/GRM-Llama3-8B-rewardmodel-ft", "Ray2333/GRM-Gemma2-2B-sftreg", "Ray2333/GRM-llama3.2-3B-sftreg" ]

[]

[ "allenai/reward-bench" ]

[ "Ray2333/reward-model-Mistral-7B-instruct-Unified-Feedback", "Ray2333/GRM-llama3-8B-distill", "Ray2333/GRM-llama3-8B-sftreg", "Ray2333/GRM-Gemma-2B-sftreg", "Ray2333/GRM-gemma2-2B-rewardmodel-ft", "Ray2333/GRM-Llama3.2-3B-rewardmodel-ft", "Ray2333/GRM-Gemma-2B-rewardmodel-ft", "Ray2333/GRM-Llama3-8B-rewardmodel-ft", "Ray2333/GRM-Gemma2-2B-sftreg", "Ray2333/GRM-llama3.2-3B-sftreg" ]

[]

[ "allenai/reward-bench" ]

1

poster

null

https://openreview.net/forum?id=juJl2uSq4D

@inproceedings{ kwon2024rl, title={{RL} in Latent {MDP}s is Tractable: Online Guarantees via Off-Policy Evaluation}, author={Jeongyeol Kwon and Shie Mannor and Constantine Caramanis and Yonathan Efroni}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=juJl2uSq4D} }

In many real-world decision problems there is partially observed, hidden or latent information that remains fixed throughout an interaction. Such decision problems can be modeled as Latent Markov Decision Processes (LMDPs), where a latent variable is selected at the beginning of an interaction and is not disclosed to the agent initially. In last decade, there has been significant progress in designing learning algorithms for solving LMDPs under different structural assumptions. However, for general LMDPs, there is no known learning algorithm that provably matches the existing lower bound. We effectively resolve this open question, introducing the first sample-efficient algorithm for LMDPs without *any additional structural assumptions*. Our result builds off a new perspective on the role off-policy evaluation guarantees and coverage coefficient in LMDPs, a perspective, which has been overlooked in the context of exploration in partially observed environments. Specifically, we establish a novel off-policy evaluation lemma and introduce a new coverage coefficient for LMDPs. Then, we show how these can be used to derive near-optimal guarantees of an optimistic exploration algorithm. These results, we believe, can be valuable for a wide range of interactive learning problems beyond the LMDP class, and especially, for partially observed environments.

RL in Latent MDPs is Tractable: Online Guarantees via Off-Policy Evaluation

[ "Jeongyeol Kwon", "Shie Mannor", "Constantine Caramanis", "Yonathan Efroni" ]

NeurIPS.cc/2024/Conference

2406.01389

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jsgYYXaSiS

@inproceedings{ zhang2024dual, title={Dual Prototype Evolving for Test-Time Generalization of Vision-Language Models}, author={Ce Zhang and Simon Stepputtis and Katia P. Sycara and Yaqi Xie}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jsgYYXaSiS} }

Test-time adaptation, which enables models to generalize to diverse data with unlabeled test samples, holds significant value in real-world scenarios. Recently, researchers have applied this setting to advanced pre-trained vision-language models (VLMs), developing approaches such as test-time prompt tuning to further extend their practical applicability. However, these methods typically focus solely on adapting VLMs from a single modality and fail to accumulate task-specific knowledge as more samples are processed. To address this, we introduce Dual Prototype Evolving (DPE), a novel test-time adaptation approach for VLMs that effectively accumulates task-specific knowledge from multi-modalities. Specifically, we create and evolve two sets of prototypes—textual and visual—to progressively capture more accurate multi-modal representations for target classes during test time. Moreover, to promote consistent multi-modal representations, we introduce and optimize learnable residuals for each test sample to align the prototypes from both modalities. Extensive experimental results on 15 benchmark datasets demonstrate that our proposed DPE consistently outperforms previous state-of-the-art methods while also exhibiting competitive computational efficiency.

Dual Prototype Evolving for Test-Time Generalization of Vision-Language Models

[ "Ce Zhang", "Simon Stepputtis", "Katia P. Sycara", "Yaqi Xie" ]

NeurIPS.cc/2024/Conference

2410.12790

[ "https://github.com/zhangce01/DPE-CLIP" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=js74ZCddxG

@inproceedings{ mai2024rflpa, title={{RFLPA}: A Robust Federated Learning Framework against Poisoning Attacks with Secure Aggregation}, author={Peihua Mai and Ran Yan and Yan Pang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=js74ZCddxG} }

Federated learning (FL) allows multiple devices to train a model collaboratively without sharing their data. Despite its benefits, FL is vulnerable to privacy leakage and poisoning attacks. To address the privacy concern, secure aggregation (SecAgg) is often used to obtain the aggregation of gradients on sever without inspecting individual user updates. Unfortunately, existing defense strategies against poisoning attacks rely on the analysis of local updates in plaintext, making them incompatible with SecAgg. To reconcile the conflicts, we propose a robust federated learning framework against poisoning attacks (RFLPA) based on SecAgg protocol. Our framework computes the cosine similarity between local updates and server updates to conduct robust aggregation. Furthermore, we leverage verifiable packed Shamir secret sharing to achieve reduced communication cost of $O(M+N)$ per user, and design a novel dot product aggregation algorithm to resolve the issue of increased information leakage. Our experimental results show that RFLPA significantly reduces communication and computation overhead by over $75\%$ compared to the state-of-the-art secret sharing method, BREA, while maintaining competitive accuracy.

RFLPA: A Robust Federated Learning Framework against Poisoning Attacks with Secure Aggregation

[ "Peihua Mai", "Ran Yan", "Yan Pang" ]

NeurIPS.cc/2024/Conference

2405.15182

[ "https://github.com/nusioraprivacy/rflpa" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=js5vZtyoIQ

@inproceedings{ yuan2024egode, title={{EGODE}: An Event-attended Graph {ODE} Framework for Modeling Rigid Dynamics}, author={Jingyang Yuan and Gongbo Sun and Zhiping Xiao and Hang Zhou and Xiao Luo and Junyu Luo and Yusheng Zhao and Wei Ju and Ming Zhang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=js5vZtyoIQ} }

This paper studies the problem of rigid dynamics modeling, which has a wide range of applications in robotics, graphics, and mechanical design. The problem is partly solved by graph neural network (GNN) simulators. However, these approaches cannot effectively handle the relationship between intrinsic continuity and instantaneous changes in rigid dynamics. Moreover, they usually neglect hierarchical structures across mesh nodes and objects in systems. In this paper, we propose a novel approach named Event-attend Graph ODE (EGODE) for effective rigid dynamics modeling. In particular, we describe the rigid system using both mesh node representations and object representations. To model continuous dynamics across hierarchical structures, we use a coupled graph ODE framework for the evolution of both types of representations over a long period. In addition, to capture instantaneous changes during the collision, we introduce an event module, which can effectively estimate the occurrence of the collision and update the states of both mesh node and object representations during evolution. Extensive experiments on a range of benchmark datasets validate the superiority of the proposed EGODE compared to various state-of-the-art baselines. The source code can be found at https://github.com/yuanjypku/EGODE.

EGODE: An Event-attended Graph ODE Framework for Modeling Rigid Dynamics

[ "Jingyang Yuan", "Gongbo Sun", "Zhiping Xiao", "Hang Zhou", "Xiao Luo", "Junyu Luo", "Yusheng Zhao", "Wei Ju", "Ming Zhang" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jrVoZLF20h

@inproceedings{ ashkenazi2024towards, title={Towards Croppable Implicit Neural Representations}, author={Maor Ashkenazi and Eran Treister}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jrVoZLF20h} }

Implicit Neural Representations (INRs) have peaked interest in recent years due to their ability to encode natural signals using neural networks. While INRs allow for useful applications such as interpolating new coordinates and signal compression, their black-box nature makes it difficult to modify them post-training. In this paper we explore the idea of editable INRs, and specifically focus on the widely used cropping operation. To this end, we present Local-Global SIRENs - a novel INR architecture that supports cropping by design. Local-Global SIRENs are based on combining local and global feature extraction for signal encoding. What makes their design unique is the ability to effortlessly remove specific portions of an encoded signal, with a proportional weight decrease. This is achieved by eliminating the corresponding weights from the network, without the need for retraining. We further show how this architecture can be used to support the straightforward extension of previously encoded signals. Beyond signal editing, we examine how the Local-Global approach can accelerate training, enhance encoding of various signals, improve downstream performance, and be applied to modern INRs such as INCODE, highlighting its potential and flexibility. Code is available at https://github.com/maorash/Local-Global-INRs.

Towards Croppable Implicit Neural Representations

[ "Maor Ashkenazi", "Eran Treister" ]

NeurIPS.cc/2024/Conference

2409.19472

[ "https://github.com/maorash/local-global-inrs" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jrNlWfor7q

@inproceedings{ dangel2024kroneckerfactored, title={Kronecker-Factored Approximate Curvature for Physics-Informed Neural Networks}, author={Felix Dangel and Johannes M{\"u}ller and Marius Zeinhofer}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jrNlWfor7q} }

Physics-Informed Neural Networks (PINNs) are infamous for being hard to train. Recently, second-order methods based on natural gradient and Gauss-Newton methods have shown promising performance, improving the accuracy achieved by first-order methods by several orders of magnitude. While promising, the proposed methods only scale to networks with a few thousand parameters due to the high computational cost to evaluate, store, and invert the curvature matrix. We propose Kronecker-factored approximate curvature (KFAC) for PINN losses that greatly reduces the computational cost and allows scaling to much larger networks. Our approach goes beyond the popular KFAC for traditional deep learning problems as it captures contributions from a PDE's differential operator that are crucial for optimization. To establish KFAC for such losses, we use Taylor-mode automatic differentiation to describe the differential operator's computation graph as a forward network with shared weights which allows us to apply a variant of KFAC for networks with weight-sharing. Empirically, we find that our KFAC-based optimizers are competitive with expensive second-order methods on small problems, scale more favorably to higher-dimensional neural networks and PDEs, and consistently outperform first-order methods.

Kronecker-Factored Approximate Curvature for Physics-Informed Neural Networks

[ "Felix Dangel", "Johannes Müller", "Marius Zeinhofer" ]

NeurIPS.cc/2024/Conference

2405.15603

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jps9KkuSD3

@inproceedings{ amoukou2024sequential, title={Sequential Harmful Shift Detection Without Labels}, author={Salim I. Amoukou and Tom Bewley and Saumitra Mishra and Freddy Lecue and Daniele Magazzeni and Manuela Veloso}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jps9KkuSD3} }

We introduce a novel approach for detecting distribution shifts that negatively impact the performance of machine learning models in continuous production environments, which requires no access to ground truth data labels. It builds upon the work of Podkopaev and Ramdas [2022], who address scenarios where labels are available for tracking model errors over time. Our solution extends this framework to work in the absence of labels, by employing a proxy for the true error. This proxy is derived using the predictions of a trained error estimator. Experiments show that our method has high power and false alarm control under various distribution shifts, including covariate and label shifts and natural shifts over geography and time.

Sequential Harmful Shift Detection Without Labels

[ "Salim I. Amoukou", "Tom Bewley", "Saumitra Mishra", "Freddy Lecue", "Daniele Magazzeni", "Manuela Veloso" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=joNPMCzVIi

@inproceedings{ guan2024improved, title={Improved Bayes Regret Bounds for Multi-Task Hierarchical Bayesian Bandit Algorithms}, author={Jiechao Guan and Hui Xiong}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=joNPMCzVIi} }

Hierarchical Bayesian bandit refers to the multi-task bandit problem in which bandit tasks are assumed to be drawn from the same distribution. In this work, we provide improved Bayes regret bounds for hierarchical Bayesian bandit algorithms in the multi-task linear bandit and semi-bandit settings. For the multi-task linear bandit, we first analyze the preexisting hierarchical Thompson sampling (HierTS) algorithm, and improve its gap-independent Bayes regret bound from $O(m\sqrt{n\log{n}\log{(mn)}})$ to $O(m\sqrt{n\log{n}})$ in the case of infinite action set, with $m$ being the number of tasks and $n$ the number of iterations per task. In the case of finite action set, we propose a novel hierarchical Bayesian bandit algorithm, named hierarchical BayesUCB (HierBayesUCB), that achieves the logarithmic but gap-dependent regret bound $O(m\log{(mn)}\log{n})$ under mild assumptions. All of the above regret bounds hold in many variants of hierarchical Bayesian linear bandit problem, including when the tasks are solved sequentially or concurrently. Furthermore, we extend the aforementioned HierTS and HierBayesUCB algorithms to the multi-task combinatorial semi-bandit setting. Concretely, our combinatorial HierTS algorithm attains comparable Bayes regret bound $O(m\sqrt{n}\log{n})$ with respect to the latest one. Moreover, our combinatorial HierBayesUCB yields a sharper Bayes regret bound $O(m\log{(mn)}\log{n})$. Experiments are conducted to validate the soundness of our theoretical results for multi-task bandit algorithms.

Improved Bayes Regret Bounds for Multi-Task Hierarchical Bayesian Bandit Algorithms

[ "Jiechao Guan", "Hui Xiong" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jjcY92FX4R

@inproceedings{ ma2024a, title={A Canonicalization Perspective on Invariant and Equivariant Learning}, author={George Ma and Yifei Wang and Derek Lim and Stefanie Jegelka and Yisen Wang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jjcY92FX4R} }

In many applications, we desire neural networks to exhibit invariance or equivariance to certain groups due to symmetries inherent in the data. Recently, frame-averaging methods emerged to be a unified framework for attaining symmetries efficiently by averaging over input-dependent subsets of the group, i.e., frames. What we currently lack is a principled understanding of the design of frames. In this work, we introduce a canonicalization perspective that provides an essential and complete view of the design of frames. Canonicalization is a classic approach for attaining invariance by mapping inputs to their canonical forms. We show that there exists an inherent connection between frames and canonical forms. Leveraging this connection, we can efficiently compare the complexity of frames as well as determine the optimality of certain frames. Guided by this principle, we design novel frames for eigenvectors that are strictly superior to existing methods --- some are even optimal --- both theoretically and empirically. The reduction to the canonicalization perspective further uncovers equivalences between previous methods. These observations suggest that canonicalization provides a fundamental understanding of existing frame-averaging methods and unifies existing equivariant and invariant learning methods. Code is available at https://github.com/PKU-ML/canonicalization.

A Canonicalization Perspective on Invariant and Equivariant Learning

[ "George Ma", "Yifei Wang", "Derek Lim", "Stefanie Jegelka", "Yisen Wang" ]

NeurIPS.cc/2024/Conference

2405.18378

[ "https://github.com/georgemlp/canonicalization" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jij4vOVU7i

@inproceedings{ wang2024efficient, title={Efficient Temporal Action Segmentation via Boundary-aware Query Voting}, author={Peiyao Wang and Yuewei Lin and Erik Blasch and Jie Wei and Haibin Ling}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jij4vOVU7i} }

Although the performance of Temporal Action Segmentation (TAS) has been improved in recent years, achieving promising results often comes with a high computational cost due to dense inputs, complex model structures, and resource-intensive post-processing requirements. To improve the efficiency while keeping the high performance, we present a novel perspective centered on per-segment classification. By harnessing the capabilities of Transformers, we tokenize each video segment as an instance token, endowed with intrinsic instance segmentation. To realize efficient action segmentation, we introduce BaFormer, a boundary-aware Transformer network. It employs instance queries for instance segmentation and a global query for class-agnostic boundary prediction, yielding continuous segment proposals. During inference, BaFormer employs a simple yet effective voting strategy to classify boundary-wise segments based on instance segmentation. Remarkably, as a single-stage approach, BaFormer significantly reduces the computational costs, utilizing only 6% of the running time compared to the state-of-the-art method DiffAct, while producing better or comparable accuracy over several popular benchmarks. The code for this project is publicly available at https://github.com/peiyao-w/BaFormer.

Efficient Temporal Action Segmentation via Boundary-aware Query Voting

[ "Peiyao Wang", "Yuewei Lin", "Erik Blasch", "Jie Wei", "Haibin Ling" ]

NeurIPS.cc/2024/Conference

2405.15995

[ "https://github.com/peiyao-w/baformer" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jgpWXnXdME

@inproceedings{ zakariaei2024advection, title={Advection Augmented Convolutional Neural Networks}, author={Niloufar Zakariaei and Siddharth Rout and Eldad Haber and Moshe Eliasof}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jgpWXnXdME} }

Many problems in physical sciences are characterized by the prediction of space-time sequences. Such problems range from weather prediction to the analysis of disease propagation and video prediction. Modern techniques for the solution of these problems typically combine Convolution Neural Networks (CNN) architecture with a time prediction mechanism. However, oftentimes, such approaches underperform in the long-range propagation of information and lack explainability. In this work, we introduce a physically inspired architecture for the solution of such problems. Namely, we propose to augment CNNs with advection by designing a novel semi-Lagrangian push operator. We show that the proposed operator allows for the non-local transformation of information compared with standard convolutional kernels. We then complement it with Reaction and Diffusion neural components to form a network that mimics the Reaction-Advection-Diffusion equation, in high dimensions. We demonstrate the effectiveness of our network on a number of spatio-temporal datasets that show their merit. Our code is available at https://github.com/Siddharth-Rout/deepADRnet.

Advection Augmented Convolutional Neural Networks

[ "Niloufar Zakariaei", "Siddharth Rout", "Eldad Haber", "Moshe Eliasof" ]

NeurIPS.cc/2024/Conference

2406.19253

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jgkKroLxeC

@inproceedings{ zhuo2024unified, title={Unified Graph Augmentations for Generalized Contrastive Learning on Graphs}, author={Jiaming Zhuo and Yintong Lu and Hui Ning and Kun Fu and Bingxin Niu and Dongxiao He and Chuan Wang and Yuanfang Guo and Zhen Wang and Xiaochun Cao and Liang Yang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jgkKroLxeC} }

In real-world scenarios, networks (graphs) and their tasks possess unique characteristics, requiring the development of a versatile graph augmentation (GA) to meet the varied demands of network analysis. Unfortunately, most Graph Contrastive Learning (GCL) frameworks are hampered by the specificity, complexity, and incompleteness of their GA techniques. Firstly, GAs designed for specific scenarios may compromise the universality of models if mishandled. Secondly, the process of identifying and generating optimal augmentations generally involves substantial computational overhead. Thirdly, the effectiveness of the GCL, even the learnable ones, is constrained by the finite selection of GAs available. To overcome the above limitations, this paper introduces a novel unified GA module dubbed UGA after reinterpreting the mechanism of GAs in GCLs from a message-passing perspective. Theoretically, this module is capable of unifying any explicit GAs, including node, edge, attribute, and subgraph augmentations. Based on the proposed UGA, a novel generalized GCL framework dubbed Graph cOntrastive UnifieD Augmentations (GOUDA) is proposed. It seamlessly integrates widely adopted contrastive losses and an introduced independence loss to fulfill the common requirements of consistency and diversity of augmentation across diverse scenarios. Evaluations across various datasets and tasks demonstrate the generality and efficiency of the proposed GOUDA over existing state-of-the-art GCLs.

Unified Graph Augmentations for Generalized Contrastive Learning on Graphs

[ "Jiaming Zhuo", "Yintong Lu", "Hui Ning", "Kun Fu", "Bingxin Niu", "Dongxiao He", "Chuan Wang", "Yuanfang Guo", "Zhen Wang", "Xiaochun Cao", "Liang Yang" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jfkid2HwNr

@inproceedings{ wang2024medformer, title={Medformer: A Multi-Granularity Patching Transformer for Medical Time-Series Classification}, author={Yihe Wang and Nan Huang and Taida Li and Yujun Yan and Xiang Zhang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jfkid2HwNr} }

Medical time series (MedTS) data, such as Electroencephalography (EEG) and Electrocardiography (ECG), play a crucial role in healthcare, such as diagnosing brain and heart diseases. Existing methods for MedTS classification primarily rely on handcrafted biomarkers extraction and CNN-based models, with limited exploration of transformer-based models. In this paper, we introduce Medformer, a multi-granularity patching transformer tailored specifically for MedTS classification. Our method incorporates three novel mechanisms to leverage the unique characteristics of MedTS: cross-channel patching to leverage inter-channel correlations, multi-granularity embedding for capturing features at different scales, and two-stage (intra- and inter-granularity) multi-granularity self-attention for learning features and correlations within and among granularities. We conduct extensive experiments on five public datasets under both subject-dependent and challenging subject-independent setups. Results demonstrate Medformer's superiority over 10 baselines, achieving top averaged ranking across five datasets on all six evaluation metrics. These findings underscore the significant impact of our method on healthcare applications, such as diagnosing Myocardial Infarction, Alzheimer's, and Parkinson's disease. We release the source code at https://github.com/DL4mHealth/Medformer.

Medformer: A Multi-Granularity Patching Transformer for Medical Time-Series Classification

[ "Yihe Wang", "Nan Huang", "Taida Li", "Yujun Yan", "Xiang Zhang" ]

NeurIPS.cc/2024/Conference

2405.19363

[ "https://github.com/dl4mhealth/medformer" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jfHkAEgKwH

@inproceedings{ wan2024locca, title={LocCa: Visual Pretraining with Location-aware Captioners}, author={Bo Wan and Michael Tschannen and Yongqin Xian and Filip Pavetic and Ibrahim Alabdulmohsin and Xiao Wang and Andr{\'e} Susano Pinto and Andreas Peter Steiner and Lucas Beyer and Xiaohua Zhai}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jfHkAEgKwH} }

Image captioning was recently found to be an effective pretraining method similar to contrastive pretraining. This opens up the largely-unexplored potential of using natural language as a flexible and powerful interface for handling diverse pretraining tasks. In this paper, we demonstrate this with a novel visual pretraining paradigm, LocCa, that incorporates location-aware tasks into captioners to teach models to extract rich information from images. Specifically, LocCa employs two tasks, bounding box prediction and location-dependent captioning, conditioned on the image pixel input. Thanks to the multitask capabilities of an encoder-decoder architecture, we show that an image captioner can effortlessly handle multiple tasks during pretraining. LocCa significantly outperforms standard captioners on downstream localization tasks, achieving state-of-the-art results on RefCOCO/+/g, while maintaining comparable performance on holistic tasks. Our work paves the way for further exploration of natural language interfaces in visual pretraining.

LocCa: Visual Pretraining with Location-aware Captioners

[ "Bo Wan", "Michael Tschannen", "Yongqin Xian", "Filip Pavetic", "Ibrahim Alabdulmohsin", "Xiao Wang", "André Susano Pinto", "Andreas Peter Steiner", "Lucas Beyer", "Xiaohua Zhai" ]

NeurIPS.cc/2024/Conference

2403.19596

[ "https://github.com/google-research/big_vision" ]

https://huggingface.co/papers/2403.19596

0

10

[]

1

poster

null

https://openreview.net/forum?id=jfE7XCE89y

@inproceedings{ han2024fusemoe, title={FuseMoE: Mixture-of-Experts Transformers for Fleximodal Fusion}, author={Xing Han and Huy Nguyen and Carl William Harris and Nhat Ho and Suchi Saria}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jfE7XCE89y} }

As machine learning models in critical fields increasingly grapple with multimodal data, they face the dual challenges of handling a wide array of modalities, often incomplete due to missing elements, and the temporal irregularity and sparsity of collected samples. Successfully leveraging this complex data, while overcoming the scarcity of high-quality training samples, is key to improving these models' predictive performance. We introduce ``FuseMoE'', a mixture-of-experts framework incorporated with an innovative gating function. Designed to integrate a diverse number of modalities, FuseMoE is effective in managing scenarios with missing modalities and irregularly sampled data trajectories. Theoretically, our unique gating function contributes to enhanced convergence rates, leading to better performance in multiple downstream tasks. The practical utility of FuseMoE in the real world is validated by a diverse set of challenging prediction tasks.

FuseMoE: Mixture-of-Experts Transformers for Fleximodal Fusion

[ "Xing Han", "Huy Nguyen", "Carl William Harris", "Nhat Ho", "Suchi Saria" ]

NeurIPS.cc/2024/Conference

2402.03226

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jeWZStUavo

@inproceedings{ xue2024reinforcement, title={Reinforcement Learning Policy as Macro Regulator Rather than Macro Placer}, author={Ke Xue and Ruo-Tong Chen and Xi Lin and Yunqi Shi and Shixiong Kai and Siyuan Xu and Chao Qian}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jeWZStUavo} }

In modern chip design, placement aims at placing millions of circuit modules, which is an essential step that significantly influences power, performance, and area (PPA) metrics. Recently, reinforcement learning (RL) has emerged as a promising technique for improving placement quality, especially macro placement. However, current RL-based placement methods suffer from long training times, low generalization ability, and inability to guarantee PPA results. A key issue lies in the problem formulation, i.e., using RL to place from scratch, which results in limits useful information and inaccurate rewards during the training process. In this work, we propose an approach that utilizes RL for the refinement stage, which allows the RL policy to learn how to adjust existing placement layouts, thereby receiving sufficient information for the policy to act and obtain relatively dense and precise rewards. Additionally, we introduce the concept of regularity during training, which is considered an important metric in the chip design industry but is often overlooked in current RL placement methods. We evaluate our approach on the ISPD 2005 and ICCAD 2015 benchmark, comparing the global half-perimeter wirelength and regularity of our proposed method against several competitive approaches. Besides, we test the PPA performance using commercial software, showing that RL as a regulator can achieve significant PPA improvements. Our RL regulator can fine-tune placements from any method and enhance their quality. Our work opens up new possibilities for the application of RL in placement, providing a more effective and efficient approach to optimizing chip design. Our code is available at \url{https://github.com/lamda-bbo/macro-regulator}.

Reinforcement Learning Policy as Macro Regulator Rather than Macro Placer

[ "Ke Xue", "Ruo-Tong Chen", "Xi Lin", "Yunqi Shi", "Shixiong Kai", "Siyuan Xu", "Chao Qian" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jdCMwF06c6

@inproceedings{ xu2024ltdefense, title={{LT}-Defense: Searching-free Backdoor Defense via Exploiting the Long-tailed Effect}, author={Yixiao Xu and Binxing Fang and Mohan Li and Keke Tang and Zhihong Tian}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jdCMwF06c6} }

Language models have shown vulnerability against backdoor attacks, threatening the security of services based on them. To mitigate the threat, existing solutions attempted to search for backdoor triggers, which can be time-consuming when handling a large search space. Looking into the attack process, we observe that poisoned data will create a long-tailed effect in the victim model, causing the decision boundary to shift towards the attack targets. Inspired by this observation, we introduce LT-Defense, the first searching-free backdoor defense via exploiting the long-tailed effect. Specifically, LT-Defense employs a small set of clean examples and two metrics to distinguish backdoor-related features in the target model. Upon detecting a backdoor model, LT-Defense additionally provides test-time backdoor freezing and attack target prediction. Extensive experiments demonstrate the effectiveness of LT-Defense in both detection accuracy and efficiency, e.g., in task-agnostic scenarios, LT-Defense achieves 98% accuracy across 1440 models with less than 1% of the time cost of state-of-the-art solutions.

LT-Defense: Searching-free Backdoor Defense via Exploiting the Long-tailed Effect

[ "Yixiao Xu", "Binxing Fang", "Mohan Li", "Keke Tang", "Zhihong Tian" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jd3msHMtTL

@inproceedings{ bardenet2024small, title={Small coresets via negative dependence: {DPP}s, linear statistics, and concentration}, author={R{\'e}mi Bardenet and Subhroshekhar Ghosh and Hugo Simon-Onfroy and Hoang-Son Tran}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jd3msHMtTL} }

Determinantal point processes (DPPs) are random configurations of points with tunable negative dependence. Because sampling is tractable, DPPs are natural candidates for subsampling tasks, such as minibatch selection or coreset construction. A \emph{coreset} is a subset of a (large) training set, such that minimizing an empirical loss averaged over the coreset is a controlled replacement for the intractable minimization of the original empirical loss. Typically, the control takes the form of a guarantee that the average loss over the coreset approximates the total loss uniformly across the parameter space. Recent work has provided significant empirical support in favor of using DPPs to build randomized coresets, coupled with interesting theoretical results that are suggestive but leave some key questions unanswered. In particular, the central question of whether the cardinality of a DPP-based coreset is fundamentally smaller than one based on independent sampling remained open. In this paper, we answer this question in the affirmative, demonstrating that \emph{DPPs can provably outperform independently drawn coresets}. In this vein, we contribute a conceptual understanding of coreset loss as a \emph{linear statistic} of the (random) coreset. We leverage this structural observation to connect the coresets problem to a more general problem of concentration phenomena for linear statistics of DPPs, wherein we obtain \emph{effective concentration inequalities that extend well-beyond the state-of-the-art}, encompassing general non-projection, even non-symmetric kernels. The latter have been recently shown to be of interest in machine learning beyond coresets, but come with a limited theoretical toolbox, to the extension of which our result contributes. Finally, we are also able to address the coresets problem for vector-valued objective functions, a novelty in the coresets literature.

Small coresets via negative dependence: DPPs, linear statistics, and concentration

[ "Rémi Bardenet", "Subhroshekhar Ghosh", "Hugo Simon-Onfroy", "Hoang-Son Tran" ]

NeurIPS.cc/2024/Conference

2411.00611

[ "https://github.com/hsimonfroy/dppcoresets" ]

-1

[]

0

oral

null

https://openreview.net/forum?id=jb5qN3212b

@inproceedings{ ma2024revisiting, title={Revisiting Score Propagation in Graph Out-of-Distribution Detection}, author={Longfei Ma and Yiyou Sun and Kaize Ding and Zemin Liu and Fei Wu}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jb5qN3212b} }

The field of graph learning has been substantially advanced by the development of deep learning models, in particular graph neural networks. However, one salient yet largely under-explored challenge is detecting Out-of-Distribution (OOD) nodes on graphs. Prevailing OOD detection techniques developed in other domains like computer vision, do not cater to the interconnected nature of graphs. This work aims to fill this gap by exploring the potential of a simple yet effective method -- OOD score propagation, which propagates OOD scores among neighboring nodes along the graph structure. This post hoc solution can be easily integrated with existing OOD scoring functions, showcasing its excellent flexibility and effectiveness in most scenarios. However, the conditions under which score propagation proves beneficial remain not fully elucidated. Our study meticulously derives these conditions and, inspired by this discovery, introduces an innovative edge augmentation strategy with theoretical guarantee. Empirical evaluations affirm the superiority of our proposed method, outperforming strong OOD detection baselines in various scenarios and settings.

Revisiting Score Propagation in Graph Out-of-Distribution Detection

[ "Longfei Ma", "Yiyou Sun", "Kaize Ding", "Zemin Liu", "Fei Wu" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ja20BpFAPa

@inproceedings{ wang2024dcgaussian, title={{DC}-Gaussian: Improving 3D Gaussian Splatting for Reflective Dash Cam Videos}, author={Linhan Wang and Kai Cheng and Shuo Lei and Shengkun Wang and Wei Yin and Chenyang Lei and Xiaoxiao Long and Chang-Tien Lu}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ja20BpFAPa} }

We present DC-Gaussian, a new method for generating novel views from in-vehicle dash cam videos. While neural rendering techniques have made significant strides in driving scenarios, existing methods are primarily designed for videos collected by autonomous vehicles. However, these videos are limited in both quantity and diversity compared to dash cam videos, which are more widely used across various types of vehicles and capture a broader range of scenarios. Dash cam videos often suffer from severe obstructions such as reflections and occlusions on the windshields, which significantly impede the application of neural rendering techniques. To address this challenge, we develop DC-Gaussian based on the recent real-time neural rendering technique 3D Gaussian Splatting (3DGS). Our approach includes an adaptive image decomposition module to model reflections and occlusions in a unified manner. Additionally, we introduce illumination-aware obstruction modeling to manage reflections and occlusions under varying lighting conditions. Lastly, we employ a geometry-guided Gaussian enhancement strategy to improve rendering details by incorporating additional geometry priors. Experiments on self-captured and public dash cam videos show that our method not only achieves state-of-the-art performance in novel view synthesis, but also accurately reconstructing captured scenes getting rid of obstructions.

DC-Gaussian: Improving 3D Gaussian Splatting for Reflective Dash Cam Videos

[ "Linhan Wang", "Kai Cheng", "Shuo Lei", "Shengkun Wang", "Wei Yin", "Chenyang Lei", "Xiaoxiao Long", "Chang-Tien Lu" ]

NeurIPS.cc/2024/Conference

2405.17705

[ "https://github.com/linhanwang/DC-Gaussian" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jZv9A8Tg9p

@inproceedings{ sun2024datafaithful, title={Data-faithful Feature Attribution: Mitigating Unobservable Confounders via Instrumental Variables}, author={Qiheng Sun and Haocheng Xia and Jinfei Liu}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jZv9A8Tg9p} }

The state-of-the-art feature attribution methods often neglect the influence of unobservable confounders, posing a risk of misinterpretation, especially when it is crucial for the interpretation to remain faithful to the data. To counteract this, we propose a new approach, data-faithful feature attribution, which trains a confounder-free model using instrumental variables. The cluttered effects of unobservable confounders in a model trained as such are decoupled from input features, thereby aligning the output of the model with the contribution of input features to the target feature in the data generation. Furthermore, feature attribution results produced by our method are more robust when focusing on attributions from the perspective of data generation. Our experiments on both synthetic and real-world datasets demonstrate the effectiveness of our approaches.

Data-faithful Feature Attribution: Mitigating Unobservable Confounders via Instrumental Variables

[ "Qiheng Sun", "Haocheng Xia", "Jinfei Liu" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jYypS5VIPj

@inproceedings{ zhang2024bridge, title={Bridge the Points: Graph-based Few-shot Segment Anything Semantically}, author={Anqi Zhang and Guangyu Gao and Jianbo Jiao and Chi Harold Liu and Yunchao Wei}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jYypS5VIPj} }

The recent advancements in large-scale pre-training techniques have significantly enhanced the capabilities of vision foundation models, notably the Segment Anything Model (SAM), which can generate precise masks based on point and box prompts. Recent studies extend SAM to Few-shot Semantic Segmentation (FSS), focusing on prompt generation for SAM-based automatic semantic segmentation. However, these methods struggle with selecting suitable prompts, require specific hyperparameter settings for different scenarios, and experience prolonged one-shot inference times due to the overuse of SAM, resulting in low efficiency and limited automation ability. To address these issues, we propose a simple yet effective approach based on graph analysis. In particular, a Positive-Negative Alignment module dynamically selects the point prompts for generating masks, especially uncovering the potential of the background context as the negative reference. Another subsequent Point-Mask Clustering module aligns the granularity of masks and selected points as a directed graph, based on mask coverage over points. These points are then aggregated by decomposing the weakly connected components of the directed graph in an efficient manner, constructing distinct natural clusters. Finally, the positive and overshooting gating, benefiting from graph-based granularity alignment, aggregates high-confident masks and filters the false-positive masks for final prediction, reducing the usage of additional hyperparameters and redundant mask generation. Extensive experimental analysis across standard FSS, One-shot Part Segmentation, and Cross Domain FSS datasets validate the effectiveness and efficiency of the proposed approach, surpassing state-of-the-art generalist models with a mIoU of 58.7% on COCO-20i and 35.2% on LVIS-92i. The project page of this work is https://andyzaq.github.io/GF-SAM/.

Bridge the Points: Graph-based Few-shot Segment Anything Semantically

[ "Anqi Zhang", "Guangyu Gao", "Jianbo Jiao", "Chi Harold Liu", "Yunchao Wei" ]

NeurIPS.cc/2024/Conference

2410.06964

[ "" ]

-1

[]

0

oral

null

https://openreview.net/forum?id=jY4PhQibmg

@inproceedings{ zhang2024gated, title={Gated Slot Attention for Efficient Linear-Time Sequence Modeling}, author={Yu Zhang and Songlin Yang and Rui-Jie Zhu and Yue Zhang and Leyang Cui and Yiqiao Wang and Bolun Wang and Freda Shi and Bailin Wang and Wei Bi and Peng Zhou and Guohong Fu}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jY4PhQibmg} }

Linear attention Transformers and their gated variants, celebrated for enabling parallel training and efficient recurrent inference, still fall short in recall-intensive tasks compared to traditional Transformers and demand significant resources for training from scratch. This paper introduces Gated Slot Attention (GSA), which enhances Attention with Bounded-memory-Control (ABC) by incorporating a gating mechanism inspired by Gated Linear Attention (GLA). Essentially, GSA comprises a two-layer GLA linked via $\operatorname{softmax}$, utilizing context-aware memory reading and adaptive forgetting to improve memory capacity while maintaining compact recurrent state size. This design greatly enhances both training and inference efficiency through GLA's hardware-efficient training algorithm and reduced state size. Additionally, retaining the $\operatorname{softmax}$ operation is particularly beneficial in ``finetuning pretrained Transformers to RNNs'' (T2R) settings, reducing the need for extensive training from scratch. Extensive experiments confirm GSA's superior performance in scenarios requiring in-context recall and in T2R settings.

Gated Slot Attention for Efficient Linear-Time Sequence Modeling

[ "Yu Zhang", "Songlin Yang", "Rui-Jie Zhu", "Yue Zhang", "Leyang Cui", "Yiqiao Wang", "Bolun Wang", "Freda Shi", "Bailin Wang", "Wei Bi", "Peng Zhou", "Guohong Fu" ]

NeurIPS.cc/2024/Conference

2409.07146

[ "https://github.com/sustcsonglin/flash-linear-attention" ]

https://huggingface.co/papers/2409.07146

5

19

2

12

[ "fla-hub/gsa-1.3B-100B", "fla-hub/gsa-2.7B-100B" ]

[]

[ "fla-hub/gsa-1.3B-100B", "fla-hub/gsa-2.7B-100B" ]

[]

1

poster

null

https://openreview.net/forum?id=jXxvSkb9HD

@inproceedings{ jansen2024statistical, title={Statistical Multicriteria Benchmarking via the {GSD}-Front}, author={Christoph Jansen and Georg Schollmeyer and Julian Rodemann and Hannah Blocher and Thomas Augustin}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jXxvSkb9HD} }

Given the vast number of classifiers that have been (and continue to be) proposed, reliable methods for comparing them are becoming increasingly important. The desire for reliability is broken down into three main aspects: (1) Comparisons should allow for different quality metrics simultaneously. (2) Comparisons should take into account the statistical uncertainty induced by the choice of benchmark suite. (3) The robustness of the comparisons under small deviations in the underlying assumptions should be verifiable. To address (1), we propose to compare classifiers using a generalized stochastic dominance ordering (GSD) and present the GSD-front as an information-efficient alternative to the classical Pareto-front. For (2), we propose a consistent statistical estimator for the GSD-front and construct a statistical test for whether a (potentially new) classifier lies in the GSD-front of a set of state-of-the-art classifiers. For (3), we relax our proposed test using techniques from robust statistics and imprecise probabilities. We illustrate our concepts on the benchmark suite PMLB and on the platform OpenML.

Statistical Multicriteria Benchmarking via the GSD-Front

[ "Christoph Jansen", "Georg Schollmeyer", "Julian Rodemann", "Hannah Blocher", "Thomas Augustin" ]

NeurIPS.cc/2024/Conference

2406.03924

[ "" ]

-1

[]

0

oral

null

https://openreview.net/forum?id=jXsxGt80sv

@inproceedings{ zhou2024staragents, title={Star-Agents: Automatic Data Optimization with {LLM} Agents for Instruction Tuning}, author={Hang Zhou and Yehui Tang and Haochen Qin and Yujie Yang and Renren Jin and Deyi Xiong and Kai Han and Yunhe Wang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jXsxGt80sv} }

The efficacy of large language models (LLMs) on downstream tasks usually hinges on instruction tuning, which relies critically on the quality of training data. Unfortunately, collecting high-quality and diverse data is both expensive and time-consuming. To mitigate this issue, we propose a novel Star-Agents framework, which automates the enhancement of data quality across datasets through multi-agent collaboration and assessment. The framework adopts a three-pronged strategy. It initially generates diverse instruction data with multiple LLM agents through a bespoke sampling method. Subsequently, the generated data undergo a rigorous evaluation using a dual-model method that assesses both difficulty and quality. Finaly, the above process evolves in a dynamic refinement phase, where more effective LLMs are prioritized, enhancing the overall data quality. Our empirical studies, including instruction tuning experiments with models such as Pythia and LLaMA, demonstrate the effectiveness of the proposed framework. Optimized datasets have achieved substantial improvements, with an average increase of 12\% and notable gains in specific metrics, such as a 40\% improvement in Fermi, as evidenced by benchmarks like MT-bench, Vicuna bench, and WizardLM testset. Codes will be released soon.

Star-Agents: Automatic Data Optimization with LLM Agents for Instruction Tuning

[ "Hang Zhou", "Yehui Tang", "Haochen Qin", "Yujie Yang", "Renren Jin", "Deyi Xiong", "Kai Han", "Yunhe Wang" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jXs6Cvpe7k

@inproceedings{ zhou2024robust, title={Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks}, author={Andy Zhou and Bo Li and Haohan Wang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jXs6Cvpe7k} }

Despite advances in AI alignment, large language models (LLMs) remain vulnerable to adversarial attacks or jailbreaking, in which adversaries can modify prompts to induce unwanted behavior. While some defenses have been proposed, they have not been adapted to newly proposed attacks and more challenging threat models. To address this, we propose an optimization-based objective for defending LLMs against jailbreaking attacks and an algorithm, Robust Prompt Optimization (RPO), to create robust system-level defenses. Our approach directly incorporates the adversary into the defensive objective and optimizes a lightweight and transferable suffix, enabling RPO to adapt to worst-case adaptive attacks. Our theoretical and experimental results show improved robustness to both jailbreaks seen during optimization and unknown jailbreaks, reducing the attack success rate (ASR) on GPT-4 to 6% and Llama-2 to 0% on JailbreakBench, setting the state-of-the-art.

Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks

[ "Andy Zhou", "Bo Li", "Haohan Wang" ]

NeurIPS.cc/2024/Conference

2401.17263

[ "https://github.com/lapisrocks/rpo" ]

https://huggingface.co/papers/2401.17263

2

1

0

3

[]

[ "TrustSafeAI/Defensive-Prompt-Patch-Jailbreak-Defense" ]

[]

[ "TrustSafeAI/Defensive-Prompt-Patch-Jailbreak-Defense" ]

1

oral

null

https://openreview.net/forum?id=jXgHEwtXs8

@inproceedings{ meng2024highresolution, title={High-Resolution Image Harmonization with Adaptive-Interval Color Transformation}, author={Quanling Meng and Qinglin Liu and Zonglin Li and Xiangyuan Lan and Shengping Zhang and Liqiang Nie}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jXgHEwtXs8} }

Existing high-resolution image harmonization methods typically rely on global color adjustments or the upsampling of parameter maps. However, these methods ignore local variations, leading to inharmonious appearances. To address this problem, we propose an Adaptive-Interval Color Transformation method (AICT), which predicts pixel-wise color transformations and adaptively adjusts the sampling interval to model local non-linearities of the color transformation at high resolution. Specifically, a parameter network is first designed to generate multiple position-dependent 3-dimensional lookup tables (3D LUTs), which use the color and position of each pixel to perform pixel-wise color transformations. Then, to enhance local variations adaptively, we separate a color transform into a cascade of sub-transformations using two 3D LUTs to achieve the non-uniform sampling intervals of the color transform. Finally, a global consistent weight learning method is proposed to predict an image-level weight for each color transform, utilizing global information to enhance the overall harmony. Extensive experiments demonstrate that our AICT achieves state-of-the-art performance with a lightweight architecture. The code is available at https://github.com/aipixel/AICT.

High-Resolution Image Harmonization with Adaptive-Interval Color Transformation

[ "Quanling Meng", "Qinglin Liu", "Zonglin Li", "Xiangyuan Lan", "Shengping Zhang", "Liqiang Nie" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jWaXhCYTV1

@inproceedings{ chen2024identifying, title={Identifying General Mechanism Shifts in Linear Causal Representations}, author={Tianyu Chen and Kevin Bello and Francesco Locatello and Bryon Aragam and Pradeep Kumar Ravikumar}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jWaXhCYTV1} }

We consider the linear causal representation learning setting where we observe a linear mixing of $d$ unknown latent factors, which follow a linear structural causal model. Recent work has shown that it is possible to recover the latent factors as well as the underlying structural causal model over them, up to permutation and scaling, provided that we have at least $d$ environments, each of which corresponds to perfect interventions on a single latent node (factor). After this powerful result, a key open problem faced by the community has been to relax these conditions: allow for coarser than perfect single-node interventions, and allow for fewer than $d$ of them, since the number of latent factors $d$ could be very large. In this work, we consider precisely such a setting, where we allow a smaller than $d$ number of environments, and also allow for very coarse interventions that can very coarsely \textit{change the entire causal graph over the latent factors}. On the flip side, we relax what we wish to extract to simply the \textit{list of nodes that have shifted between one or more environments}. We provide a surprising identifiability result that it is indeed possible, under some very mild standard assumptions, to identify the set of shifted nodes. Our identifiability proof moreover is a constructive one: we explicitly provide necessary and sufficient conditions for a node to be a shifted node, and show that we can check these conditions given observed data. Our algorithm lends itself very naturally to the sample setting where instead of just interventional distributions, we are provided datasets of samples from each of these distributions. We corroborate our results on both synthetic experiments as well as an interesting psychometric dataset. The code can be found at https://github.com/TianyuCodings/iLCS.

Identifying General Mechanism Shifts in Linear Causal Representations

[ "Tianyu Chen", "Kevin Bello", "Francesco Locatello", "Bryon Aragam", "Pradeep Kumar Ravikumar" ]

NeurIPS.cc/2024/Conference

2410.24059

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jWGGEDYORs

@inproceedings{ yan2024darnet, title={{DARN}et: Dual Attention Refinement Network with Spatiotemporal Construction for Auditory Attention Detection}, author={Sheng Yan and Cunhang Fan and Hongyu Zhang and Xiaoke Yang and Jianhua Tao and Zhao Lv}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jWGGEDYORs} }

At a cocktail party, humans exhibit an impressive ability to direct their attention. The auditory attention detection (AAD) approach seeks to identify the attended speaker by analyzing brain signals, such as EEG signals. However, current AAD algorithms overlook the spatial distribution information within EEG signals and lack the ability to capture long-range latent dependencies, limiting the model's ability to decode brain activity. To address these issues, this paper proposes a dual attention refinement network with spatiotemporal construction for AAD, named DARNet, which consists of the spatiotemporal construction module, dual attention refinement module, and feature fusion \& classifier module. Specifically, the spatiotemporal construction module aims to construct more expressive spatiotemporal feature representations, by capturing the spatial distribution characteristics of EEG signals. The dual attention refinement module aims to extract different levels of temporal patterns in EEG signals and enhance the model's ability to capture long-range latent dependencies. The feature fusion \& classifier module aims to aggregate temporal patterns and dependencies from different levels and obtain the final classification results. The experimental results indicate that DARNet achieved excellent classification performance, particularly under short decision windows. While maintaining excellent classification performance, DARNet significantly reduces the number of required parameters. Compared to the state-of-the-art models, DARNet reduces the parameter count by 91\%. Code is available at: https://github.com/fchest/DARNet.git.

DARNet: Dual Attention Refinement Network with Spatiotemporal Construction for Auditory Attention Detection

[ "Sheng Yan", "Cunhang Fan", "Hongyu Zhang", "Xiaoke Yang", "Jianhua Tao", "Zhao Lv" ]

NeurIPS.cc/2024/Conference

2410.11181

[ "https://github.com/fchest/darnet" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jV6z08u7y0

@inproceedings{ wang2024the, title={The Implicit Bias of Gradient Descent toward Collaboration between Layers: A Dynamic Analysis of Multilayer Perceptions}, author={Zheng Wang and Geyong Min and Wenjie Ruan}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jV6z08u7y0} }

The implicit bias of gradient descent has long been considered the primary mechanism explaining the superior generalization of over-parameterized neural networks without overfitting, even when the training error is zero. However, the implicit bias toward adversarial robustness has rarely been considered in the research community, although it is crucial for the trustworthiness of machine learning models. To fill this gap, in this paper, we explore whether consecutive layers collaborate to strengthen adversarial robustness during gradient descent. By quantifying this collaboration between layers using our proposed concept, co-correlation, we demonstrate a monotonically increasing trend in co-correlation, which implies a decreasing trend in adversarial robustness during gradient descent. Additionally, we observe different behaviours between narrow and wide neural networks during gradient descent. We conducted extensive experiments that verified our proposed theorems.

The Implicit Bias of Gradient Descent toward Collaboration between Layers: A Dynamic Analysis of Multilayer Perceptions

[ "Zheng Wang", "Geyong Min", "Wenjie Ruan" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jURBh4V9N4

@inproceedings{ weiminbai2024an, title={An Expectation-Maximization Algorithm for Training Clean Diffusion Models from Corrupted Observations}, author={WeiminBai and Yifei Wang and Wenzheng Chen and He Sun}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jURBh4V9N4} }

Diffusion models excel in solving imaging inverse problems due to their ability to model complex image priors. However, their reliance on large, clean datasets for training limits their practical use where clean data is scarce. In this paper, we propose EMDiffusion, an expectation-maximization (EM) approach to train diffusion models from corrupted observations. Our method alternates between reconstructing clean images from corrupted data using a known diffusion model (E-step) and refining diffusion model weights based on these reconstructions (M-step). This iterative process leads the learned diffusion model to gradually converge to a local optimum, that is, to approximate the true clean data distribution. We validate our method through extensive experiments on diverse computational imaging tasks, including random inpainting, denoising, and deblurring, achieving new state-of-the-art performance.

An Expectation-Maximization Algorithm for Training Clean Diffusion Models from Corrupted Observations

[ "WeiminBai", "Yifei Wang", "Wenzheng Chen", "He Sun" ]

NeurIPS.cc/2024/Conference

2407.01014

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jTyjwRpLZ5

@inproceedings{ yu2024stochastic, title={Stochastic Zeroth-Order Optimization under Strongly Convexity and Lipschitz Hessian: Minimax Sample Complexity}, author={Qian Yu and Yining Wang and Baihe Huang and Qi Lei and Jason D. Lee}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jTyjwRpLZ5} }

Optimization of convex functions under stochastic zeroth-order feedback has been a major and challenging question in online learning. In this work, we consider the problem of optimizing second-order smooth and strongly convex functions where the algorithm is only accessible to noisy evaluations of the objective function it queries. We provide the first tight characterization for the rate of the minimax simple regret by developing matching upper and lower bounds. We propose an algorithm that features a combination of a bootstrapping stage and a mirror-descent stage. Our main technical innovation consists of a sharp characterization for the spherical-sampling gradient estimator under higher-order smoothness conditions, which allows the algorithm to optimally balance the bias-variance tradeoff, and a new iterative method for the bootstrapping stage, which maintains the performance for unbounded Hessian.

Stochastic Zeroth-Order Optimization under Strongly Convexity and Lipschitz Hessian: Minimax Sample Complexity

[ "Qian Yu", "Yining Wang", "Baihe Huang", "Qi Lei", "Jason D. Lee" ]

NeurIPS.cc/2024/Conference

2406.19617

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jS34QpqdWs

@inproceedings{ sun2024nonstationary, title={Nonstationary Sparse Spectral Permanental Process}, author={Zicheng Sun and Yixuan Zhang and Zenan Ling and Xuhui Fan and Feng Zhou}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jS34QpqdWs} }

Existing permanental processes often impose constraints on kernel types or stationarity, limiting the model's expressiveness. To overcome these limitations, we propose a novel approach utilizing the sparse spectral representation of nonstationary kernels. This technique relaxes the constraints on kernel types and stationarity, allowing for more flexible modeling while reducing computational complexity to the linear level. Additionally, we introduce a deep kernel variant by hierarchically stacking multiple spectral feature mappings, further enhancing the model's expressiveness to capture complex patterns in data. Experimental results on both synthetic and real-world datasets demonstrate the effectiveness of our approach, particularly in scenarios with pronounced data nonstationarity. Additionally, ablation studies are conducted to provide insights into the impact of various hyperparameters on model performance.

Nonstationary Sparse Spectral Permanental Process

[ "Zicheng Sun", "Yixuan Zhang", "Zenan Ling", "Xuhui Fan", "Feng Zhou" ]

NeurIPS.cc/2024/Conference

2410.03581

[ "https://github.com/SZC20/DNSSPP" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jRtxzzk0a6

@inproceedings{ prabhakar2024kraken, title={Kraken: Inherently Parallel Transformers For Efficient Multi-Device Inference}, author={Rohan Baskar Prabhakar and Hengrui Zhang and David Wentzlaff}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jRtxzzk0a6} }

Large Transformer networks are increasingly used in settings where low inference latency is necessary to enable new applications and improve the end-user experience. However, autoregressive inference is resource intensive and requires parallelism for efficiency. Parallelism introduces collective communication that is both expensive and represents a phase when hardware resources are underutilized. Towards mitigating this, Kraken is an evolution of the standard Transformer architecture that is designed to complement existing tensor parallelism schemes for efficient inference on multi-device systems. By introducing a fixed degree of intra-layer model parallelism, the architecture allows collective operations to be overlapped with compute, decreasing latency and increasing hardware utilization. When trained on OpenWebText, Kraken models reach a similar perplexity as standard Transformers while also preserving their language modeling capabilities as evaluated on the SuperGLUE benchmark. Importantly, when tested on multi-GPU systems using TensorRT-LLM engines, Kraken speeds up Time To First Token by a mean of 35.6% across a range of model sizes, context lengths, and degrees of tensor parallelism.

Kraken: Inherently Parallel Transformers For Efficient Multi-Device Inference

[ "Rohan Baskar Prabhakar", "Hengrui Zhang", "David Wentzlaff" ]

NeurIPS.cc/2024/Conference

2408.07802

[ "" ]

https://huggingface.co/papers/2408.07802

0

3

[]

1

poster

null

https://openreview.net/forum?id=jMJVFP4BH6

@inproceedings{ fang2024towards, title={Towards Neuron Attributions in Multi-Modal Large Language Models}, author={Junfeng Fang and Zac Bi and Ruipeng Wang and Houcheng Jiang and Yuan Gao and Kun Wang and An Zhang and Jie Shi and Xiang Wang and Tat-Seng Chua}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jMJVFP4BH6} }

As Large Language Models (LLMs) demonstrate impressive capabilities, demystifying their internal mechanisms becomes increasingly vital. Neuron attribution, which attributes LLM outputs to specific neurons to reveal the semantic properties they learn, has emerged as a key interpretability approach. However, while neuron attribution has made significant progress in deciphering text-only LLMs, its application to Multimodal LLMs (MLLMs) remains less explored. To address this gap, we propose a novel Neuron Attribution method tailored for MLLMs, termed NAM. Specifically, NAM not only reveals the modality-specific semantic knowledge learned by neurons within MLLMs, but also highlights several intriguing properties of neurons, such as cross-modal invariance and semantic sensitivity. These properties collectively elucidate the inner workings mechanism of MLLMs, providing a deeper understanding of how MLLMs process and generate multi-modal content. Through theoretical analysis and empirical validation, we demonstrate the efficacy of NAM and the valuable insights it offers. Furthermore, leveraging NAM, we introduce a multi-modal knowledge editing paradigm, underscoring the practical significance of our approach for downstream applications of MLLMs.

Towards Neuron Attributions in Multi-Modal Large Language Models

[ "Junfeng Fang", "Zac Bi", "Ruipeng Wang", "Houcheng Jiang", "Yuan Gao", "Kun Wang", "An Zhang", "Jie Shi", "Xiang Wang", "Tat-Seng Chua" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jM9atrvUii

@inproceedings{ groth2024kermut, title={Kermut: Composite kernel regression for protein variant effects}, author={Peter M{\o}rch Groth and Mads Herbert Kerrn and Lars Olsen and Jesper Salomon and Wouter Boomsma}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jM9atrvUii} }

Reliable prediction of protein variant effects is crucial for both protein optimization and for advancing biological understanding. For practical use in protein engineering, it is important that we can also provide reliable uncertainty estimates for our predictions, and while prediction accuracy has seen much progress in recent years, uncertainty metrics are rarely reported. We here provide a Gaussian process regression model, Kermut, with a novel composite kernel for modeling mutation similarity, which obtains state-of-the-art performance for supervised protein variant effect prediction while also offering estimates of uncertainty through its posterior. An analysis of the quality of the uncertainty estimates demonstrates that our model provides meaningful levels of overall calibration, but that instance-specific uncertainty calibration remains more challenging.

Kermut: Composite kernel regression for protein variant effects

[ "Peter Mørch Groth", "Mads Herbert Kerrn", "Lars Olsen", "Jesper Salomon", "Wouter Boomsma" ]

NeurIPS.cc/2024/Conference

2407.00002

[ "https://github.com/petergroth/kermut" ]

-1

[]

0

oral

null

https://openreview.net/forum?id=jLUbLxa4XV

@inproceedings{ rekavandi2024certified, title={Certified Adversarial Robustness via Randomized \${\textbackslash}alpha\$-Smoothing for Regression Models}, author={Aref Miri Rekavandi and Farhad Farokhi and Olga Ohrimenko and Benjamin I. P. Rubinstein}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jLUbLxa4XV} }

Certified adversarial robustness of large-scale deep networks has progressed substantially after the introduction of randomized smoothing. Deep net classifiers are now provably robust in their predictions against a large class of threat models, including $\ell_1$, $\ell_2$, and $\ell_\infty$ norm-bounded attacks. Certified robustness analysis by randomized smoothing has not been performed for deep regression networks where the output variable is continuous and unbounded. In this paper, we extend the existing results for randomized smoothing into regression models using powerful tools from robust statistics, in particular, $\alpha$-trimming filter as the smoothing function. Adjusting the hyperparameter $\alpha$ achieves a smooth trade-off between desired certified robustness and utility. For the first time, we propose a benchmark for certified robust regression in visual positioning systems using the Cambridge Landmarks dataset where robustness analysis is essential for autonomous navigation of AI agents and self-driving cars. Code is publicly available at \url{https://github.com/arekavandi/Certified_adv_RRegression/}.

Certified Adversarial Robustness via Randomized α-Smoothing for Regression Models

[ "Aref Miri Rekavandi", "Farhad Farokhi", "Olga Ohrimenko", "Benjamin I. P. Rubinstein" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jL0EsbfbAV

@inproceedings{ wang2024exploring, title={Exploring Behavior-Relevant and Disentangled Neural Dynamics with Generative Diffusion Models}, author={Yule Wang and Chengrui Li and Weihan Li and Anqi Wu}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jL0EsbfbAV} }

Understanding the neural basis of behavior is a fundamental goal in neuroscience. Current research in large-scale neuro-behavioral data analysis often relies on decoding models, which quantify behavioral information in neural data but lack details on behavior encoding. This raises an intriguing scientific question: "how can we enable in-depth exploration of neural representations in behavioral tasks, revealing interpretable neural dynamics associated with behaviors". However, addressing this issue is challenging due to the varied behavioral encoding across different brain regions and mixed selectivity at the population level. To tackle this limitation, our approach, named ("BeNeDiff"), first identifies a fine-grained and disentangled neural subspace using a behavior-informed latent variable model. It then employs state-of-the-art generative diffusion models to synthesize behavior videos that interpret the neural dynamics of each latent factor. We validate the method on multi-session datasets containing widefield calcium imaging recordings across the dorsal cortex. Through guiding the diffusion model to activate individual latent factors, we verify that the neural dynamics of latent factors in the disentangled neural subspace provide interpretable quantifications of the behaviors of interest. At the same time, the neural subspace in BeNeDiff demonstrates high disentanglement and neural reconstruction quality.

Exploring Behavior-Relevant and Disentangled Neural Dynamics with Generative Diffusion Models

[ "Yule Wang", "Chengrui Li", "Weihan Li", "Anqi Wu" ]

NeurIPS.cc/2024/Conference

2410.09614

[ "https://github.com/yulewang97/benediff" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jKLyKeZfzv

@inproceedings{ zhang2024motenas, title={{MOTE}-{NAS}: Multi-Objective Training-based Estimate for Efficient Neural Architecture Search}, author={Yuming Zhang and Jun Wei Hsieh and Xin Li and Ming-Ching Chang and Chun-Chieh Lee and Kuo-Chin Fan}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jKLyKeZfzv} }

Neural Architecture Search (NAS) methods seek effective optimization toward performance metrics regarding model accuracy and generalization while facing challenges regarding search costs and GPU resources. Recent Neural Tangent Kernel (NTK) NAS methods achieve remarkable search efficiency based on a training-free model estimate; however, they overlook the non-convex nature of the DNNs in the search process. In this paper, we develop Multi-Objective Training-based Estimate (MOTE) for efficient NAS, retaining search effectiveness and achieving the new state-of-the-art in the accuracy and cost trade-off. To improve NTK and inspired by the Training Speed Estimation (TSE) method, MOTE is designed to model the actual performance of DNNs from macro to micro perspective by draw loss landscape and convergence speed simultaneously. Using two reduction strategies, the MOTE is generated based on a reduced architecture and a reduced dataset. Inspired by evolutionary search, our iterative ranking-based, coarse-to-fine architecture search is highly effective. Experiments on NASBench-201 show MOTE-NAS achieves 94.32% accuracy on CIFAR-10, 72.81% on CIFAR-100, and 46.38% on ImageNet-16-120, outperforming NTK-based NAS approaches. An evaluation-free (EF) version of MOTE-NAS delivers high efficiency in only 5 minutes, delivering a model more accurate than KNAS.

MOTE-NAS: Multi-Objective Training-based Estimate for Efficient Neural Architecture Search

[ "Yuming Zhang", "Jun Wei Hsieh", "Xin Li", "Ming-Ching Chang", "Chun-Chieh Lee", "Kuo-Chin Fan" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jImXgQEmX3

@inproceedings{ guan2024amor, title={{AMOR}: A Recipe for Building Adaptable Modular Knowledge Agents Through Process Feedback}, author={Jian Guan and Wei Wu and zujie wen and Peng Xu and Hongning Wang and Minlie Huang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jImXgQEmX3} }

The notable success of large language models (LLMs) has sparked an upsurge in building language agents to complete various complex tasks. We present AMOR, an agent framework based on open-source LLMs, which reasons with external knowledge bases and adapts to specific domains through human supervision to the reasoning process. AMOR builds reasoning logic over a finite state machine (FSM) that solves problems through autonomous executions and transitions over disentangled modules. This allows humans to provide direct feedback to the individual modules, and thus naturally forms process supervision. Based on this reasoning and feedback framework, we develop AMOR through two-stage fine-tuning: warm-up and adaptation. The former fine-tunes the LLM with examples automatically constructed from various public datasets, enabling AMOR to generalize across different knowledge environments, while the latter tailors AMOR to specific domains using process feedback. Extensive experiments across multiple domains demonstrate the advantage of AMOR to strong baselines, thanks to its FSM-based reasoning and process feedback mechanism. The code and data are publicly available at https://github.com/JianGuanTHU/AMOR.

AMOR: A Recipe for Building Adaptable Modular Knowledge Agents Through Process Feedback

[ "Jian Guan", "Wei Wu", "zujie wen", "Peng Xu", "Hongning Wang", "Minlie Huang" ]

NeurIPS.cc/2024/Conference

2402.01469

[ "https://github.com/jianguanthu/amor" ]

https://huggingface.co/papers/2402.01469

0

6

[ "Jiann/AMOR-warmup", "Jiann/AMOR-adaptation" ]

[ "Jiann/AMOR_warmup_data" ]

[]

[ "Jiann/AMOR-warmup", "Jiann/AMOR-adaptation" ]

[ "Jiann/AMOR_warmup_data" ]

[]

1

poster

null

https://openreview.net/forum?id=jIh4W7r0rn

@inproceedings{ luo2024autoregressive, title={Autoregressive Image Diffusion: Generation of Image Sequence and Application in {MRI}}, author={Guanxiong Luo and Shoujin Huang and Martin Uecker}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jIh4W7r0rn} }

Magnetic resonance imaging (MRI) is a widely used non-invasive imaging modality. However, a persistent challenge lies in balancing image quality with imaging speed. This trade-off is primarily constrained by k-space measurements, which traverse specific trajectories in the spatial Fourier domain (k-space). These measurements are often undersampled to shorten acquisition times, resulting in image artifacts and compromised quality. Generative models learn image distributions and can be used to reconstruct high-quality images from undersampled k-space data. In this work, we present the autoregressive image diffusion (AID) model for image sequences and use it to sample the posterior for accelerated MRI reconstruction. The algorithm incorporates both undersampled k-space and pre-existing information. Models trained with fastMRI dataset are evaluated comprehensively. The results show that the AID model can robustly generate sequentially coherent image sequences. In MRI applications, the AID can outperform the standard diffusion model and reduce hallucinations, due to the learned inter-image dependencies. The project code is available at https://github.com/mrirecon/aid.

Autoregressive Image Diffusion: Generation of Image Sequence and Application in MRI

[ "Guanxiong Luo", "Shoujin Huang", "Martin Uecker" ]

NeurIPS.cc/2024/Conference

2405.14327

[ "https://github.com/mrirecon/aid" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jIabKyXOTt

@inproceedings{ jin2024sparsityagnostic, title={Sparsity-Agnostic Linear Bandits with Adaptive Adversaries}, author={Tianyuan Jin and Kyoungseok Jang and Nicol{\`o} Cesa-Bianchi}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jIabKyXOTt} }

We study stochastic linear bandits where, in each round, the learner receives a set of actions (i.e., feature vectors), from which it chooses an element and obtains a stochastic reward. The expected reward is a fixed but unknown linear function of the chosen action. We study \emph{sparse} regret bounds, that depend on the number $S$ of non-zero coefficients in the linear reward function. Previous works focused on the case where $S$ is known, or the action sets satisfy additional assumptions. In this work, we obtain the first sparse regret bounds that hold when $S$ is unknown and the action sets are adversarially generated. Our techniques combine online to confidence set conversions with a novel randomized model selection approach over a hierarchy of nested confidence sets. When $S$ is known, our analysis recovers state-of-the-art bounds for adversarial action sets. We also show that a variant of our approach, using Exp3 to dynamically select the confidence sets, can be used to improve the empirical performance of stochastic linear bandits while enjoying a regret bound with optimal dependence on the time horizon.

Sparsity-Agnostic Linear Bandits with Adaptive Adversaries

[ "Tianyuan Jin", "Kyoungseok Jang", "Nicolò Cesa-Bianchi" ]

NeurIPS.cc/2024/Conference

2406.01192

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jHh804fZ5l

@inproceedings{ sakaue2024generalization, title={Generalization Bound and Learning Methods for Data-Driven Projections in Linear Programming}, author={Shinsaku Sakaue and Taihei Oki}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jHh804fZ5l} }

How to solve high-dimensional linear programs (LPs) efficiently is a fundamental question. Recently, there has been a surge of interest in reducing LP sizes using *random projections*, which can accelerate solving LPs independently of improving LP solvers. This paper explores a new direction of *data-driven projections*, which use projection matrices learned from data instead of random projection matrices. Given training data of $n$-dimensional LPs, we learn an $n\times k$ projection matrix with $n > k$. When addressing a future LP instance, we reduce its dimensionality from $n$ to $k$ via the learned projection matrix, solve the resulting LP to obtain a $k$-dimensional solution, and apply the learned matrix to it to recover an $n$-dimensional solution. On the theoretical side, a natural question is: how much data is sufficient to ensure the quality of recovered solutions? We address this question based on the framework of *data-driven algorithm design*, which connects the amount of data sufficient for establishing generalization bounds to the *pseudo-dimension* of performance metrics. We obtain an $\tilde{\mathrm{O}}(nk^2)$ upper bound on the pseudo-dimension, where $\tilde{\mathrm{O}}$ compresses logarithmic factors. We also provide an $\Omega(nk)$ lower bound, implying our result is tight up to an $\tilde{\mathrm{O}}(k)$ factor. On the practical side, we explore two simple methods for learning projection matrices: PCA- and gradient-based methods. While the former is relatively efficient, the latter can sometimes achieve better solution quality. Experiments demonstrate that learning projection matrices from data is indeed beneficial: it leads to significantly higher solution quality than the existing random projection while greatly reducing the time for solving LPs.

Generalization Bound and Learning Methods for Data-Driven Projections in Linear Programming

[ "Shinsaku Sakaue", "Taihei Oki" ]

NeurIPS.cc/2024/Conference

2309.00203

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jFWl9EWZ7z

@inproceedings{ zhang2024multiobject, title={Multi-Object 3D Grounding with Dynamic Modules and Language-Informed Spatial Attention}, author={Haomeng Zhang and Chiao An Yang and Raymond A. Yeh}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jFWl9EWZ7z} }

Multi-object 3D Grounding involves locating 3D boxes based on a given query phrase from a point cloud. It is a challenging and significant task that has numerous applications in visual understanding, human-computer interaction, and robotics. To tackle this challenge, we introduce D-LISA, a two-stage approach that incorporates three innovations. First, a dynamic vision module that enables a variable and learnable number of box proposals. Second, a dynamic camera positioning that extracts features for each proposal. Third, a language-informed spatial attention module that better reasons over the proposals to output the final prediction. Empirically, experiments show that our method outperforms the state-of-the-art methods on multi-object 3D grounding by 12.8% (absolute) and is competitive in single-object 3D grounding.

Multi-Object 3D Grounding with Dynamic Modules and Language-Informed Spatial Attention

[ "Haomeng Zhang", "Chiao An Yang", "Raymond A. Yeh" ]

NeurIPS.cc/2024/Conference

2410.22306

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jDF2ZXI8AX

@inproceedings{ hong2024mvcyl, title={{MV}2Cyl: Reconstructing 3D Extrusion Cylinders from Multi-View Images}, author={Eunji Hong and Nguyen Minh Hieu and Mikaela Angelina Uy and Minhyuk Sung}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jDF2ZXI8AX} }

We present MV2Cyl, a novel method for reconstructing 3D from 2D multi-view images, not merely as a field or raw geometry but as a sketch-extrude CAD. Extracting extrusion cylinders from raw 3D geometry has been extensively researched in computer vision, while the processing of 3D data through neural networks has remained a bottleneck. Since 3D scans are generally accompanied by multi-view images, leveraging 2D convolutional neural networks allows these images to be exploited as a rich source for extracting extrusion cylinder information. However, we observe that extracting only the surface information of the extrudes and utilizing it results in suboptimal outcomes due to the challenges in the occlusion and surface segmentation. By synergizing with the extracted base curve information, we achieve the optimal reconstruction result with the best accuracy in 2D sketch and extrude parameter estimation. Our experiments, comparing our method with previous work that takes a raw 3D point cloud as input, demonstrate the effectiveness of our approach by taking advantage of multi-view images.

MV2Cyl: Reconstructing 3D Extrusion Cylinders from Multi-View Images

[ "Eunji Hong", "Nguyen Minh Hieu", "Mikaela Angelina Uy", "Minhyuk Sung" ]

NeurIPS.cc/2024/Conference

2406.10853

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jCPufQaHvb

@inproceedings{ chen2024conm, title={Con4m: Context-aware Consistency Learning Framework for Segmented Time Series Classification}, author={Junru Chen and Tianyu Cao and Jing Xu and Jiahe Li and Zhilong Chen and Tao Xiao and Yang Yang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jCPufQaHvb} }

Time Series Classification (TSC) encompasses two settings: classifying entire sequences or classifying segmented subsequences. The raw time series for segmented TSC usually contain Multiple classes with Varying Duration of each class (MVD). Therefore, the characteristics of MVD pose unique challenges for segmented TSC, yet have been largely overlooked by existing works. Specifically, there exists a natural temporal dependency between consecutive instances (segments) to be classified within MVD. However, mainstream TSC models rely on the assumption of independent and identically distributed (i.i.d.), focusing on independently modeling each segment. Additionally, annotators with varying expertise may provide inconsistent boundary labels, leading to unstable performance of noise-free TSC models. To address these challenges, we first formally demonstrate that valuable contextual information enhances the discriminative power of classification instances. Leveraging the contextual priors of MVD at both the data and label levels, we propose a novel consistency learning framework Con4m, which effectively utilizes contextual information more conducive to discriminating consecutive segments in segmented TSC tasks, while harmonizing inconsistent boundary labels for training. Extensive experiments across multiple datasets validate the effectiveness of Con4m in handling segmented TSC tasks on MVD. The source code is available at https://github.com/MrNobodyCali/Con4m.

Con4m: Context-aware Consistency Learning Framework for Segmented Time Series Classification

[ "Junru Chen", "Tianyu Cao", "Jing Xu", "Jiahe Li", "Zhilong Chen", "Tao Xiao", "Yang Yang" ]

NeurIPS.cc/2024/Conference

2408.00041

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jCMYIUwprx

@inproceedings{ le2024indict, title={{INDICT}: Code Generation with Internal Dialogues of Critiques for Both Security and Helpfulness}, author={Hung Le and Doyen Sahoo and Yingbo Zhou and Caiming Xiong and Silvio Savarese}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jCMYIUwprx} }

Large language models (LLMs) for code are typically trained to align with natural language instructions to closely follow their intentions and requirements. However, in many practical scenarios, it becomes increasingly challenging for these models to navigate the intricate boundary between helpfulness and safety, especially against highly complex yet potentially malicious instructions. In this work, we introduce INDICT: a new framework that empowers LLMs with Internal Dialogues of Critiques for both safety and helpfulness guidance. The internal dialogue is a dual cooperative system between a safety-driven critic and a helpfulness-driven critic. Each critic provides analysis against the given task and corresponding generated response, equipped with external knowledge queried through relevant code snippets and tools like web search and code interpreter. We engage the dual critic system in both code generation stage as well as code execution stage, providing preemptive and post-hoc guidance respectively to LLMs. We evaluated INDICT on 8 diverse tasks across 8 programming languages from 5 benchmarks, using LLMs from 7B to 70B parameters. We observed that our approach can provide an advanced level of critiques of both safety and helpfulness analysis, significantly improving the quality of output codes (+10% absolute improvements in all models).

INDICT: Code Generation with Internal Dialogues of Critiques for Both Security and Helpfulness

[ "Hung Le", "Doyen Sahoo", "Yingbo Zhou", "Caiming Xiong", "Silvio Savarese" ]

NeurIPS.cc/2024/Conference

2407.02518

[ "https://github.com/SalesforceAIResearch/indict_code_gen" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=jBf3eIyD2x

@inproceedings{ hayase2024querybased, title={Query-Based Adversarial Prompt Generation}, author={Jonathan Hayase and Ema Borevkovi{\'c} and Nicholas Carlini and Florian Tram{\`e}r and Milad Nasr}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=jBf3eIyD2x} }

Recent work has shown it is possible to construct adversarial examples that cause aligned language models to emit harmful strings or perform harmful behavior. Existing attacks work either in the white-box setting (with full access to the model weights), or through _transferability_: the phenomenon that adversarial examples crafted on one model often remain effective on other models. We improve on prior work with a _query-based_ attack that leverages API access to a remote language model to construct adversarial examples that cause the model to emit harmful strings with (much) higher probability than with transfer-only attacks. We validate our attack on GPT-3.5 and OpenAI's safety classifier; we can cause GPT-3.5 to emit harmful strings that current transfer attacks fail at, and we can evade the OpenAI and Llama Guard safety classifiers with nearly 100% probability.

Query-Based Adversarial Prompt Generation

[ "Jonathan Hayase", "Ema Borevković", "Nicholas Carlini", "Florian Tramèr", "Milad Nasr" ]

NeurIPS.cc/2024/Conference

2402.12329

[ "" ]

https://huggingface.co/papers/2402.12329

0

5

[]

1

poster

null

https://openreview.net/forum?id=j7sw0nXLjZ

@inproceedings{ wu2024instructorinspired, title={Instructor-inspired Machine Learning for Robust Molecular Property Prediction}, author={Fang Wu and Shuting Jin and Siyuan Li and Stan Z. Li}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=j7sw0nXLjZ} }

Machine learning catalyzes a revolution in chemical and biological science. However, its efficacy is heavily dependent on the availability of labeled data, and annotating biochemical data is extremely laborious. To surmount this data sparsity challenge, we present an instructive learning algorithm named InstructMol to measure pseudo-labels' reliability and help the target model leverage large-scale unlabeled data. InstructMol does not require transferring knowledge between multiple domains, which avoids the potential gap between the pretraining and fine-tuning stages. We demonstrated the high accuracy of InstructMol on several real-world molecular datasets and out-of-distribution (OOD) benchmarks.

Instructor-inspired Machine Learning for Robust Molecular Property Prediction

[ "Fang Wu", "Shuting Jin", "Siyuan Li", "Stan Z. Li" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=j6kJSS9O6I

@inproceedings{ qiao2024agent, title={Agent Planning with World Knowledge Model}, author={Shuofei Qiao and Runnan Fang and Ningyu Zhang and Yuqi Zhu and Xiang Chen and Shumin Deng and Yong Jiang and Pengjun Xie and Fei Huang and Huajun Chen}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=j6kJSS9O6I} }

Recent endeavors towards directly using large language models (LLMs) as agent models to execute interactive planning tasks have shown commendable results. Despite their achievements, however, they still struggle with brainless trial-and-error in global planning and generating hallucinatory actions in local planning due to their poor understanding of the "real" physical world. Imitating humans' mental world knowledge model which provides global prior knowledge before the task and maintains local dynamic knowledge during the task, in this paper, we introduce parametric World Knowledge Model (WKM) to facilitate agent planning. Concretely, we steer the agent model to self-synthesize knowledge from both expert and sampled trajectories. Then we develop WKM, providing prior task knowledge to guide the global planning and dynamic state knowledge to assist the local planning. Experimental results on three real-world simulated datasets with Mistral-7B, Gemma-7B, and Llama-3-8B demonstrate that our method can achieve superior performance compared to various strong baselines. Besides, we analyze to illustrate that our WKM can effectively alleviate the blind trial-and-error and hallucinatory action issues, providing strong support for the agent's understanding of the world. Other interesting findings include: 1) our instance-level task knowledge can generalize better to unseen tasks, 2) weak WKM can guide strong agent model planning, and 3) unified WKM training has promising potential for further development.

Agent Planning with World Knowledge Model

[ "Shuofei Qiao", "Runnan Fang", "Ningyu Zhang", "Yuqi Zhu", "Xiang Chen", "Shumin Deng", "Yong Jiang", "Pengjun Xie", "Fei Huang", "Huajun Chen" ]

NeurIPS.cc/2024/Conference

2405.14205

[ "https://github.com/zjunlp/wkm" ]

https://huggingface.co/papers/2405.14205

2

0

10

[ "zjunlp/WKM-mistral-alfworld" ]

[]

[ "zjunlp/WKM-mistral-alfworld" ]

[]

1

poster

null

https://openreview.net/forum?id=j6Zsoj544N

@inproceedings{ hu2024does, title={Does Worst-Performing Agent Lead the Pack? Analyzing Agent Dynamics in Unified Distributed {SGD}}, author={Jie Hu and Yi-Ting Ma and Do Young Eun}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=j6Zsoj544N} }

Distributed learning is essential to train machine learning algorithms across *heterogeneous* agents while maintaining data privacy. We conduct an asymptotic analysis of Unified Distributed SGD (UD-SGD), exploring a variety of communication patterns, including decentralized SGD and local SGD within Federated Learning (FL), as well as the increasing communication interval in the FL setting. In this study, we assess how different sampling strategies, such as *i.i.d.* sampling, shuffling, and Markovian sampling, affect the convergence speed of UD-SGD by considering the impact of agent dynamics on the limiting covariance matrix as described in the Central Limit Theorem (CLT). Our findings not only support existing theories on linear speedup and asymptotic network independence, but also theoretically and empirically show how efficient sampling strategies employed by individual agents contribute to overall convergence in UD-SGD. Simulations reveal that a few agents using highly efficient sampling can achieve or surpass the performance of the majority employing moderately improved strategies, providing new insights beyond traditional analyses focusing on the worst-performing agent.

Does Worst-Performing Agent Lead the Pack? Analyzing Agent Dynamics in Unified Distributed SGD

[ "Jie Hu", "Yi-Ting Ma", "Do Young Eun" ]

NeurIPS.cc/2024/Conference

2409.17499

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=j2wCrWmgMX

@inproceedings{ nikitin2024kernel, title={Kernel Language Entropy: Fine-grained Uncertainty Quantification for {LLM}s from Semantic Similarities}, author={Alexander V Nikitin and Jannik Kossen and Yarin Gal and Pekka Marttinen}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=j2wCrWmgMX} }

Uncertainty quantification in Large Language Models (LLMs) is crucial for applications where safety and reliability are important. In particular, uncertainty can be used to improve the trustworthiness of LLMs by detecting factually incorrect model responses, commonly called hallucinations. Critically, one should seek to capture the model's semantic uncertainty, i.e., the uncertainty over the meanings of LLM outputs, rather than uncertainty over lexical or syntactic variations that do not affect answer correctness. To address this problem, we propose Kernel Language Entropy (KLE), a novel method for uncertainty estimation in white- and black-box LLMs. KLE defines positive semidefinite unit trace kernels to encode the semantic similarities of LLM outputs and quantifies uncertainty using the von Neumann entropy. It considers pairwise semantic dependencies between answers (or semantic clusters), providing more fine-grained uncertainty estimates than previous methods based on hard clustering of answers. We theoretically prove that KLE generalizes the previous state-of-the-art method called semantic entropy and empirically demonstrate that it improves uncertainty quantification performance across multiple natural language generation datasets and LLM architectures.

Kernel Language Entropy: Fine-grained Uncertainty Quantification for LLMs from Semantic Similarities

[ "Alexander V Nikitin", "Jannik Kossen", "Yarin Gal", "Pekka Marttinen" ]

NeurIPS.cc/2024/Conference

2405.20003

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=j2hzCTqbF0

@inproceedings{ wu2024accurate, title={Accurate and Steady Inertial Pose Estimation through Sequence Structure Learning and Modulation}, author={Yinghao Wu and chaoran wang and Lu Yin and Shihui Guo and Yipeng Qin}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=j2hzCTqbF0} }

Transformer models excel at capturing long-range dependencies in sequential data, but lack explicit mechanisms to leverage structural patterns inherent in fixed-length input sequences. In this paper, we propose a novel sequence structure learning and modulation approach that endows Transformers with the ability to model and utilize such fixed-sequence structural properties for improved performance on inertial pose estimation tasks. Specifically, our method introduces a Sequence Structure Module (SSM) that utilizes structural information of fixed-length inertial sensor readings to adjust the input features of transformers. Such structural information can either be acquired by learning or specified based on users' prior knowledge. To justify the prospect of our approach, we show that i) injecting spatial structural information of IMUs/joints learned from data improves accuracy, while ii) injecting temporal structural information based on smooth priors reduces jitter (i.e., improves steadiness), in a spatial-temporal transformer solution for inertial pose estimation. Extensive experiments across multiple benchmark datasets demonstrate the superiority of our approach against state-of-the-art methods and has the potential to advance the design of the transformer architecture for fixed-length sequences.

Accurate and Steady Inertial Pose Estimation through Sequence Structure Learning and Modulation

[ "Yinghao Wu", "chaoran wang", "Lu Yin", "Shihui Guo", "Yipeng Qin" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=j25WK4GEGH

@inproceedings{ cho2024generalizable, title={Generalizable Person Re-identification via Balancing Alignment and Uniformity}, author={Yoonki Cho and Jaeyoon Kim and Woo Jae Kim and Junsik Jung and Sung-eui Yoon}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=j25WK4GEGH} }

Domain generalizable person re-identification (DG re-ID) aims to learn discriminative representations that are robust to distributional shifts. While data augmentation is a straightforward solution to improve generalization, certain augmentations exhibit a polarized effect in this task, enhancing in-distribution performance while deteriorating out-of-distribution performance. In this paper, we investigate this phenomenon and reveal that it leads to sparse representation spaces with reduced uniformity. To address this issue, we propose a novel framework, Balancing Alignment and Uniformity (BAU), which effectively mitigates this effect by maintaining a balance between alignment and uniformity. Specifically, BAU incorporates alignment and uniformity losses applied to both original and augmented images and integrates a weighting strategy to assess the reliability of augmented samples, further improving the alignment loss. Additionally, we introduce a domain-specific uniformity loss that promotes uniformity within each source domain, thereby enhancing the learning of domain-invariant features. Extensive experimental results demonstrate that BAU effectively exploits the advantages of data augmentation, which previous studies could not fully utilize, and achieves state-of-the-art performance without requiring complex training procedures.

Generalizable Person Re-identification via Balancing Alignment and Uniformity

[ "Yoonki Cho", "Jaeyoon Kim", "Woo Jae Kim", "Junsik Jung", "Sung-eui Yoon" ]

NeurIPS.cc/2024/Conference

2411.11471

[ "https://github.com/yoonkicho/bau" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=j14wStqZni

@inproceedings{ ullah2024publicdata, title={Public-data Assisted Private Stochastic Optimization: Power and Limitations}, author={Enayat Ullah and Michael Menart and Raef Bassily and Crist{\'o}bal A Guzm{\'a}n and Raman Arora}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=j14wStqZni} }

We study the limits and capability of public-data assisted differentially private (PA-DP) algorithms. Specifically, we focus on the problem of stochastic convex optimization (SCO) with either labeled or unlabeled public data. For complete/labeled public data, we show that any $(\epsilon,\delta)$-PA-DP has excess risk $\tilde{\Omega}\big(\min(\frac{1}{\sqrt{n_{\text{pub}}}},\frac{1}{\sqrt{n}}+\frac{\sqrt{d}}{n\epsilon} ) \big)$, where $d$ is the dimension, ${n_{\text{pub}}}$ is the number of public samples, ${n_{\text{priv}}}$ is the number of private samples, and $n={n_{\text{pub}}}+{n_{\text{priv}}}$. These lower bounds are established via our new lower bounds for PA-DP mean estimation, which are of a similar form. Up to constant factors, these lower bounds show that the simple strategy of either treating all data as private or discarding the private data, is optimal. We also study PA-DP supervised learning with \textit{unlabeled} public samples. In contrast to our previous result, we here show novel methods for leveraging public data in private supervised learning. For generalized linear models (GLM) with unlabeled public data, we show an efficient algorithm which, given $\tilde{O}({n_{\text{priv}}}\epsilon)$ unlabeled public samples, achieves the dimension independent rate $\tilde{O}\big(\frac{1}{\sqrt{{n_{\text{priv}}}}} + \frac{1}{\sqrt{{n_{\text{priv}}}\epsilon}}\big)$. We develop new lower bounds for this setting which shows that this rate cannot be improved with more public samples, and any fewer public samples leads to a worse rate. Finally, we provide extensions of this result to general hypothesis classes with finite \textit{fat-shattering dimension} with applications to neural networks and non-Euclidean geometries.

Public-data Assisted Private Stochastic Optimization: Power and Limitations

[ "Enayat Ullah", "Michael Menart", "Raef Bassily", "Cristóbal A Guzmán", "Raman Arora" ]

NeurIPS.cc/2024/Conference

2403.03856

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iykao97YXf

@inproceedings{ le2024reinforcement, title={Reinforcement Learning with {LTL} and \${\textbackslash}omega\$-Regular Objectives via Optimality-Preserving Translation to Average Rewards}, author={Xuan-Bach Le and Dominik Wagner and Leon Witzman and Alexander Rabinovich and Luke Ong}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iykao97YXf} }

Linear temporal logic (LTL) and, more generally, $\omega$-regular objectives are alternatives to the traditional discount sum and average reward objectives in reinforcement learning (RL), offering the advantage of greater comprehensibility and hence explainability. In this work, we study the relationship between these objectives. Our main result is that each RL problem for $\omega$-regular objectives can be reduced to a limit-average reward problem in an optimality-preserving fashion, via (finite-memory) reward machines. Furthermore, we demonstrate the efficacy of this approach by showing that optimal policies for limit-average problems can be found asymptotically by solving a sequence of discount-sum problems approximately. Consequently, we resolve an open problem: optimal policies for LTL and $\omega$-regular objectives can be learned asymptotically.

Reinforcement Learning with LTL and ω-Regular Objectives via Optimality-Preserving Translation to Average Rewards

[ "Xuan-Bach Le", "Dominik Wagner", "Leon Witzman", "Alexander Rabinovich", "Luke Ong" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ivCX2cjwcT

@inproceedings{ timilsina2024identifiable, title={Identifiable Shared Component Analysis of Unpaired Multimodal Mixtures}, author={Subash Timilsina and Sagar Shrestha and Xiao Fu}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ivCX2cjwcT} }

A core task in multi-modal learning is to integrate information from multiple feature spaces (e.g., text and audio), offering modality-invariant essential representations of data. Recent research showed that, classical tools such as canonical correlation analysis (CCA) provably identify the shared components up to minor ambiguities, when samples in each modality are generated from a linear mixture of shared and private components. Such identifiability results were obtained under the condition that the cross-modality samples are aligned/paired according to their shared information. This work takes a step further, investigating shared component identifiability from multi-modal linear mixtures where cross-modality samples are unaligned. A distribution divergence minimization-based loss is proposed, under which a suite of sufficient conditions ensuring identifiability of the shared components are derived. Our conditions are based on cross-modality distribution discrepancy characterization and density-preserving transform removal, which are much milder than existing studies relying on independent component analysis. More relaxed conditions are also provided via adding reasonable structural constraints, motivated by available side information in various applications. The identifiability claims are thoroughly validated using synthetic and real-world data.

Identifiable Shared Component Analysis of Unpaired Multimodal Mixtures

[ "Subash Timilsina", "Sagar Shrestha", "Xiao Fu" ]

NeurIPS.cc/2024/Conference

2409.19422

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=itztwTAcN6

@inproceedings{ mao2024a, title={A Universal Growth Rate for Learning with Smooth Surrogate Losses}, author={Anqi Mao and Mehryar Mohri and Yutao Zhong}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=itztwTAcN6} }

This paper presents a comprehensive analysis of the growth rate of $H$-consistency bounds (and excess error bounds) for various surrogate losses used in classification. We prove a square-root growth rate near zero for smooth margin-based surrogate losses in binary classification, providing both upper and lower bounds under mild assumptions. This result also translates to excess error bounds. Our lower bound requires weaker conditions than those in previous work for excess error bounds, and our upper bound is entirely novel. Moreover, we extend this analysis to multi-class classification with a series of novel results, demonstrating a universal square-root growth rate for smooth *comp-sum* and *constrained losses*, covering common choices for training neural networks in multi-class classification. Given this universal rate, we turn to the question of choosing among different surrogate losses. We first examine how $H$-consistency bounds vary across surrogates based on the number of classes. Next, ignoring constants and focusing on behavior near zero, we identify *minimizability gaps* as the key differentiating factor in these bounds. Thus, we thoroughly analyze these gaps, to guide surrogate loss selection, covering: comparisons across different comp-sum losses, conditions where gaps become zero, and general conditions leading to small gaps. Additionally, we demonstrate the key role of minimizability gaps in comparing excess error bounds and $H$-consistency bounds.

A Universal Growth Rate for Learning with Smooth Surrogate Losses

[ "Anqi Mao", "Mehryar Mohri", "Yutao Zhong" ]

NeurIPS.cc/2024/Conference

2405.05968

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=itbKmreqUZ

@inproceedings{ zhang2024cosmic, title={{COSMIC}: Compress Satellite Image Efficiently via Diffusion Compensation}, author={Ziyuan Zhang and Han Qiu and Zhang Maosen and Jun Liu and Bin Chen and Tianwei Zhang and Hewu Li}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=itbKmreqUZ} }

With the rapidly increasing number of satellites in space and their enhanced capabilities, the amount of earth observation images collected by satellites is exceeding the transmission limits of satellite-to-ground links. Although existing learned image compression solutions achieve remarkable performance by using a sophisticated encoder to extract fruitful features as compression and using a decoder to reconstruct. It is still hard to directly deploy those complex encoders on current satellites' embedded GPUs with limited computing capability and power supply to compress images in orbit. In this paper, we propose COSMIC, a simple yet effective learned compression solution to transmit satellite images. We first design a lightweight encoder (i.e. reducing FLOPs by 2.5~5X) on satellite to achieve a high image compression ratio to save satellite-to-ground links. Then, for reconstructions on the ground, to deal with the feature extraction ability degradation due to simplifying encoders, we propose a diffusion-based model to compensate image details when decoding. Our insight is that satellite's earth observation photos are not just images but indeed multi-modal data with a nature of Text-to-Image pairing since they are collected with rich sensor data (e.g. coordinates, timestep, etc.) that can be used as the condition for diffusion generation. Extensive experiments show that COSMIC outperforms state-of-the-art baselines on both perceptual and distortion metrics.

COSMIC: Compress Satellite Image Efficiently via Diffusion Compensation

[ "Ziyuan Zhang", "Han Qiu", "Zhang Maosen", "Jun Liu", "Bin Chen", "Tianwei Zhang", "Hewu Li" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=isZ8XRe3De

@inproceedings{ kong2024customizing, title={Customizing Language Models with Instance-wise Lo{RA} for Sequential Recommendation}, author={Xiaoyu Kong and Jiancan Wu and An Zhang and Leheng Sheng and Hui Lin and Xiang Wang and Xiangnan He}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=isZ8XRe3De} }

Sequential recommendation systems predict the next interaction item based on users' past interactions, aligning recommendations with individual preferences. Leveraging the strengths of Large Language Models (LLMs) in knowledge comprehension and reasoning, recent approaches are eager to apply LLMs to sequential recommendation. A common paradigm is converting user behavior sequences into instruction data, and fine-tuning the LLM with parameter-efficient fine-tuning (PEFT) methods like Low-Rank Adaption (LoRA). However, the uniform application of LoRA across diverse user behaviors is insufficient to capture individual variability, resulting in negative transfer between disparate sequences. To address these challenges, we propose Instance-wise LoRA (iLoRA). We innovatively treat the sequential recommendation task as a form of multi-task learning, integrating LoRA with the Mixture of Experts (MoE) framework. This approach encourages different experts to capture various aspects of user behavior. Additionally, we introduce a sequence representation guided gate function that generates customized expert participation weights for each user sequence, which allows dynamic parameter adjustment for instance-wise recommendations. In sequential recommendation, iLoRA achieves an average relative improvement of 11.4\% over basic LoRA in the hit ratio metric, with less than a 1\% relative increase in trainable parameters. Extensive experiments on three benchmark datasets demonstrate the effectiveness of iLoRA, highlighting its superior performance compared to existing methods in mitigating negative transfer and improving recommendation accuracy. Our data and code are available at https://github.com/AkaliKong/iLoRA.

Customizing Language Models with Instance-wise LoRA for Sequential Recommendation

[ "Xiaoyu Kong", "Jiancan Wu", "An Zhang", "Leheng Sheng", "Hui Lin", "Xiang Wang", "Xiangnan He" ]

NeurIPS.cc/2024/Conference

2408.10159

[ "https://github.com/akalikong/ilora" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ioe66JeCMF

@inproceedings{ wang2024time, title={Time Makes Space: Emergence of Place Fields in Networks Encoding Temporally Continuous Sensory Experiences}, author={Zhaoze Wang and Ronald W Di Tullio and Spencer Rooke and Vijay Balasubramanian}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ioe66JeCMF} }

The vertebrate hippocampus is thought to use recurrent connectivity in area CA3 to support episodic memory recall from partial cues. This brain area also contains place cells, whose location-selective firing fields implement maps supporting spatial memory. Here we show that place cells emerge in networks trained to remember temporally continuous sensory episodes. We model CA3 as a recurrent autoencoder that recalls and reconstructs sensory experiences from noisy and partially occluded observations by agents traversing simulated arenas. The agents move in realistic trajectories modeled from rodents and environments are modeled as continuously varying, high-dimensional, sensory experience maps (spatially smoothed Gaussian random fields). Training our autoencoder to accurately pattern-complete and reconstruct sensory experiences with a constraint on total activity causes spatially localized firing fields, i.e., place cells, to emerge in the encoding layer. The emergent place fields reproduce key aspects of hippocampal phenomenology: a) remapping (maintenance of and reversion to distinct learned maps in different environments), implemented via repositioning of experience manifolds in the network’s hidden layer, b) orthogonality of spatial representations in different arenas, c) robust place field emergence in differently shaped rooms, with single units showing multiple place fields in large or complex spaces, and (d) slow representational drift of place fields. We argue that these results arise because continuous traversal of space makes sensory experience temporally continuous. We make testable predictions: a) rapidly changing sensory context will disrupt place fields, b) place fields will form even if recurrent connections are blocked, but reversion to previously learned representations upon remapping will be abolished, c) the dimension of temporally smooth experience sets the dimensionality of place fields, including during virtual navigation of abstract spaces.

Time Makes Space: Emergence of Place Fields in Networks Encoding Temporally Continuous Sensory Experiences

[ "Zhaoze Wang", "Ronald W Di Tullio", "Spencer Rooke", "Vijay Balasubramanian" ]

NeurIPS.cc/2024/Conference

2408.05798

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ioKQzb8SMr

@inproceedings{ yun2024guided, title={Guided Trajectory Generation with Diffusion Models for Offline Model-based Optimization}, author={Taeyoung Yun and Sujin Yun and Jaewoo Lee and Jinkyoo Park}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ioKQzb8SMr} }

Optimizing complex and high-dimensional black-box functions is ubiquitous in science and engineering fields. Unfortunately, the online evaluation of these functions is restricted due to time and safety constraints in most cases. In offline model-based optimization (MBO), we aim to find a design that maximizes the target function using only a pre-existing offline dataset. While prior methods consider forward or inverse approaches to address the problem, these approaches are limited by conservatism and the difficulty of learning highly multi-modal mappings. Recently, there has been an emerging paradigm of learning to improve solutions with synthetic trajectories constructed from the offline dataset. In this paper, we introduce a novel conditional generative modeling approach to produce trajectories toward high-scoring regions. First, we construct synthetic trajectories toward high-scoring regions using the dataset while injecting locality bias for consistent improvement directions. Then, we train a conditional diffusion model to generate trajectories conditioned on their scores. Lastly, we sample multiple trajectories from the trained model with guidance to explore high-scoring regions beyond the dataset and select high-fidelity designs among generated trajectories with the proxy function. Extensive experiment results demonstrate that our method outperforms competitive baselines on Design-Bench and its practical variants. The code is publicly available in \url{https://github.com/dbsxodud-11/GTG}.

Guided Trajectory Generation with Diffusion Models for Offline Model-based Optimization

[ "Taeyoung Yun", "Sujin Yun", "Jaewoo Lee", "Jinkyoo Park" ]

NeurIPS.cc/2024/Conference

2407.01624

[ "https://github.com/dbsxodud-11/gtg" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ioAlzcELTf

@inproceedings{ xu2024generalizing, title={Generalizing Weather Forecast to Fine-grained Temporal Scales via Physics-{AI} Hybrid Modeling}, author={Wanghan Xu and Fenghua Ling and Wenlong Zhang and Tao Han and Hao Chen and Wanli Ouyang and LEI BAI}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ioAlzcELTf} }

Data-driven artificial intelligence (AI) models have made significant advancements in weather forecasting, particularly in medium-range and nowcasting. However, most data-driven weather forecasting models are black-box systems that focus on learning data mapping rather than fine-grained physical evolution in the time dimension. Consequently, the limitations in the temporal scale of datasets prevent these models from forecasting at finer time scales. This paper proposes a physics-AI hybrid model (i.e., WeatherGFT) which Generalizes weather forecasts to Finer-grained Temporal scales beyond training dataset. Specifically, we employ a carefully designed PDE kernel to simulate physical evolution on a small time scale (e.g., 300 seconds) and use a parallel neural networks with a learnable router for bias correction. Furthermore, we introduce a lead time-aware training framework to promote the generalization of the model at different lead times. The weight analysis of physics-AI modules indicates that physics conducts major evolution while AI performs corrections adaptively. Extensive experiments show that WeatherGFT trained on an hourly dataset, achieves state-of-the-art performance across multiple lead times and exhibits the capability to generalize 30-minute forecasts.

Generalizing Weather Forecast to Fine-grained Temporal Scales via Physics-AI Hybrid Modeling

[ "Wanghan Xu", "Fenghua Ling", "Wenlong Zhang", "Tao Han", "Hao Chen", "Wanli Ouyang", "LEI BAI" ]

NeurIPS.cc/2024/Conference

2405.13796

[ "https://github.com/black-yt/weathergft" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ik37kKxKBm

@inproceedings{ yang2024incontext, title={In-Context Learning with Representations: Contextual Generalization of Trained Transformers}, author={Tong Yang and Yu Huang and Yingbin Liang and Yuejie Chi}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ik37kKxKBm} }

In-context learning (ICL) refers to a remarkable capability of pretrained large language models, which can learn a new task given a few examples during inference. However, theoretical understanding of ICL is largely under-explored, particularly whether transformers can be trained to generalize to unseen examples in a prompt, which will require the model to acquire contextual knowledge of the prompt for generalization. This paper investigates the training dynamics of transformers by gradient descent through the lens of non-linear regression tasks. The contextual generalization here can be attained via learning the template function for each task in-context, where all template functions lie in a linear space with $m$ basis functions. We analyze the training dynamics of one-layer multi-head transformers to {in-contextly} predict unlabeled inputs given partially labeled prompts, where the labels contain Gaussian noise and the number of examples in each prompt are not sufficient to determine the template. Under mild assumptions, we show that the training loss for a one-layer multi-head transformer converges linearly to a global minimum. Moreover, the transformer effectively learns to perform ridge regression over the basis functions. To our knowledge, this study is the first provable demonstration that transformers can learn contextual (i.e., template) information to generalize to both unseen examples and tasks when prompts contain only a small number of query-answer pairs.

In-Context Learning with Representations: Contextual Generalization of Trained Transformers

[ "Tong Yang", "Yu Huang", "Yingbin Liang", "Yuejie Chi" ]

NeurIPS.cc/2024/Conference

2408.10147

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iiYadgKHwo

@inproceedings{ zhou2024variational, title={Variational Distillation of Diffusion Policies into Mixture of Experts}, author={Hongyi Zhou and Denis Blessing and Ge Li and Onur Celik and Xiaogang Jia and Gerhard Neumann and Rudolf Lioutikov}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iiYadgKHwo} }

This work introduces Variational Diffusion Distillation (VDD), a novel method that distills denoising diffusion policies into Mixtures of Experts (MoE) through variational inference. Diffusion Models are the current state-of-the-art in generative modeling due to their exceptional ability to accurately learn and represent complex, multi-modal distributions. This ability allows Diffusion Models to replicate the inherent diversity in human behavior, making them the preferred models in behavior learning such as Learning from Human Demonstrations (LfD). However, diffusion models come with some drawbacks, including the intractability of likelihoods and long inference times due to their iterative sampling process. The inference times, in particular, pose a significant challenge to real-time applications such as robot control. In contrast, MoEs effectively address the aforementioned issues while retaining the ability to represent complex distributions but are notoriously difficult to train. VDD is the first method that distills pre-trained diffusion models into MoE models, and hence, combines the expressiveness of Diffusion Models with the benefits of Mixture Models. Specifically, VDD leverages a decompositional upper bound of the variational objective that allows the training of each expert separately, resulting in a robust optimization scheme for MoEs. VDD demonstrates across nine complex behavior learning tasks, that it is able to: i) accurately distill complex distributions learned by the diffusion model, ii) outperform existing state-of-the-art distillation methods, and iii) surpass conventional methods for training MoE. The code and videos are available at https://intuitive-robots.github.io/vdd-website.

Variational Distillation of Diffusion Policies into Mixture of Experts

[ "Hongyi Zhou", "Denis Blessing", "Ge Li", "Onur Celik", "Xiaogang Jia", "Gerhard Neumann", "Rudolf Lioutikov" ]

NeurIPS.cc/2024/Conference

2406.12538

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ihEHCbqZEx

@inproceedings{ yun2024flexmoe, title={Flex-MoE: Modeling Arbitrary Modality Combination via the Flexible Mixture-of-Experts}, author={Sukwon Yun and Inyoung Choi and Jie Peng and Yangfan Wu and Jingxuan Bao and Qiyiwen Zhang and Jiayi Xin and Qi Long and Tianlong Chen}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ihEHCbqZEx} }

Multimodal learning has gained increasing importance across various fields, offering the ability to integrate data from diverse sources such as images, text, and personalized records, which are frequently observed in medical domains. However, in scenarios where some modalities are missing, many existing frameworks struggle to accommodate arbitrary modality combinations, often relying heavily on a single modality or complete data. This oversight of potential modality combinations limits their applicability in real-world situations. To address this challenge, we propose Flex-MoE (Flexible Mixture-of-Experts), a new framework designed to flexibly incorporate arbitrary modality combinations while maintaining robustness to missing data. The core idea of Flex-MoE is to first address missing modalities using a new missing modality bank that integrates observed modality combinations with the corresponding missing ones. This is followed by a uniquely designed Sparse MoE framework. Specifically, Flex-MoE first trains experts using samples with all modalities to inject generalized knowledge through the generalized router ($\mathcal{G}$-Router). The $\mathcal{S}$-Router then specializes in handling fewer modality combinations by assigning the top-1 gate to the expert corresponding to the observed modality combination. We evaluate Flex-MoE on the ADNI dataset, which encompasses four modalities in the Alzheimer's Disease domain, as well as on the MIMIC-IV dataset. The results demonstrate the effectiveness of Flex-MoE, highlighting its ability to model arbitrary modality combinations in diverse missing modality scenarios. Code is available at: \url{https://github.com/UNITES-Lab/flex-moe}.

Flex-MoE: Modeling Arbitrary Modality Combination via the Flexible Mixture-of-Experts

[ "Sukwon Yun", "Inyoung Choi", "Jie Peng", "Yangfan Wu", "Jingxuan Bao", "Qiyiwen Zhang", "Jiayi Xin", "Qi Long", "Tianlong Chen" ]

NeurIPS.cc/2024/Conference

2410.08245

[ "https://github.com/unites-lab/flex-moe" ]

-1

[]

0

oral

null

https://openreview.net/forum?id=ieYdf9TZ2u

@inproceedings{ zhuo2024luminanext, title={Lumina-Next : Making Lumina-T2X Stronger and Faster with Next-DiT}, author={Le Zhuo and Ruoyi Du and Han Xiao and Yangguang Li and Dongyang Liu and Rongjie Huang and Wenze Liu and Xiangyang Zhu and Fu-Yun Wang and Zhanyu Ma and Xu Luo and Zehan Wang and Kaipeng Zhang and Lirui Zhao and Si Liu and Xiangyu Yue and Wanli Ouyang and Yu Qiao and Hongsheng Li and Peng Gao}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ieYdf9TZ2u} }

Lumina-T2X is a nascent family of Flow-based Large Diffusion Transformers (Flag-DiT) that establishes a unified framework for transforming noise into various modalities, such as images and videos, conditioned on text instructions. Despite its promising capabilities, Lumina-T2X still encounters challenges including training instability, slow inference, and extrapolation artifacts. In this paper, we present Lumina-Next, an improved version of Lumina-T2X, showcasing stronger generation performance with increased training and inference efficiency. We begin with a comprehensive analysis of the Flag-DiT architecture and identify several suboptimal components, which we address by introducing the Next-DiT architecture with 3D RoPE and sandwich normalizations. To enable better resolution extrapolation, we thoroughly compare different context extrapolation methods applied to text-to-image generation with 3D RoPE, and propose Frequency- and Time-Aware Scaled RoPE tailored for diffusion transformers. Additionally, we introduce a sigmoid time discretization schedule for diffusion sampling, which achieves high-quality generation in 5-10 steps combined with higher-order ODE solvers. Thanks to these improvements, Lumina-Next not only improves the basic text-to-image generation but also demonstrates superior resolution extrapolation capabilities as well as multilingual generation using decoder-based LLMs as the text encoder, all in a zero-shot manner. To further validate Lumina-Next as a versatile generative framework, we instantiate it on diverse tasks including visual recognition, multi-views, audio, music, and point cloud generation, showcasing strong performance across these domains. By releasing all codes and model weights at https://github.com/Alpha-VLLM/Lumina-T2X, we aim to advance the development of next-generation generative AI capable of universal modeling.

Lumina-Next : Making Lumina-T2X Stronger and Faster with Next-DiT

[ "Le Zhuo", "Ruoyi Du", "Han Xiao", "Yangguang Li", "Dongyang Liu", "Rongjie Huang", "Wenze Liu", "Xiangyang Zhu", "Fu-Yun Wang", "Zhanyu Ma", "Xu Luo", "Zehan Wang", "Kaipeng Zhang", "Lirui Zhao", "Si Liu", "Xiangyu Yue", "Wanli Ouyang", "Yu Qiao", "Hongsheng Li", "Peng Gao" ]

NeurIPS.cc/2024/Conference

[ "https://github.com/alpha-vllm/lumina-t2x" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ibKpPabHVn

@inproceedings{ shen2024deepdrk, title={Deep{DRK}: Deep Dependency Regularized Knockoff for Feature Selection}, author={Hongyu Shen and Yici Yan and Zhizhen Zhao}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ibKpPabHVn} }

Model-X knockoff has garnered significant attention among various feature selection methods due to its guarantees for controlling the false discovery rate (FDR). Since its introduction in parametric design, knockoff techniques have evolved to handle arbitrary data distributions using deep learning-based generative models. However, we have observed limitations in the current implementations of the deep Model-X knockoff framework. Notably, the "swap property" that knockoffs require often faces challenges at the sample level, resulting in diminished selection power. To address these issues, we develop "Deep Dependency Regularized Knockoff (DeepDRK)," a distribution-free deep learning method that effectively balances FDR and power. In DeepDRK, we introduce a novel formulation of the knockoff model as a learning problem under multi-source adversarial attacks. By employing an innovative perturbation technique, we achieve lower FDR and higher power. Our model outperforms existing benchmarks across synthetic, semi-synthetic, and real-world datasets, particularly when sample sizes are small and data distributions are non-Gaussian.

DeepDRK: Deep Dependency Regularized Knockoff for Feature Selection

[ "Hongyu Shen", "Yici Yan", "Zhizhen Zhao" ]

NeurIPS.cc/2024/Conference

2402.17176

[ "https://github.com/nowonder2000/deepdrk" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=ia4WUCwHA9

@inproceedings{ silveri2024theoretical, title={Theoretical guarantees in {KL} for Diffusion Flow Matching}, author={Marta Gentiloni Silveri and Alain Oliviero Durmus and Giovanni Conforti}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=ia4WUCwHA9} }

Flow Matching (FM) (also referred to as stochastic interpolants or rectified flows) stands out as a class of generative models that aims to bridge in finite time the target distribution $\nu^\star$ with an auxiliary distribution $\mu$ leveraging a fixed coupling $\pi$ and a bridge which can either be deterministic or stochastic. These two ingredients define a path measure which can then be approximated by learning the drift of its Markovian projection. The main contribution of this paper is to provide relatively mild assumption on $\nu^\star$, $\mu$ and $\pi$ to obtain non-asymptotics guarantees for Diffusion Flow Matching (DFM) models using as bridge the conditional distribution associated with the Brownian motion. More precisely, it establishes bounds on the Kullback-Leibler divergence between the target distribution and the one generated by such DFM models under moment conditions on the score of $\nu^\star$, $\mu$ and $\pi$, and a standard $\mathrm{L}^2$-drift-approximation error assumption.

Theoretical guarantees in KL for Diffusion Flow Matching

[ "Marta Gentiloni Silveri", "Alain Oliviero Durmus", "Giovanni Conforti" ]

NeurIPS.cc/2024/Conference

2409.08311

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iYzyTmd3Jd

@inproceedings{ gao2024coohoi, title={Coo{HOI}: Learning Cooperative Human-Object Interaction with Manipulated Object Dynamics}, author={Jiawei Gao and Ziqin Wang and Zeqi Xiao and Jingbo Wang and Tai Wang and Jinkun Cao and Xiaolin Hu and Si Liu and Jifeng Dai and Jiangmiao Pang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iYzyTmd3Jd} }

Enabling humanoid robots to clean rooms has long been a pursued dream within humanoid research communities. However, many tasks require multi-humanoid collaboration, such as carrying large and heavy furniture together. Given the scarcity of motion capture data on multi-humanoid collaboration and the efficiency challenges associated with multi-agent learning, these tasks cannot be straightforwardly addressed using training paradigms designed for single-agent scenarios. In this paper, we introduce **Coo**perative **H**uman-**O**bject **I**nteraction (**CooHOI**), a framework designed to tackle the challenge of multi-humanoid object transportation problem through a two-phase learning paradigm: individual skill learning and subsequent policy transfer. First, a single humanoid character learns to interact with objects through imitation learning from human motion priors. Then, the humanoid learns to collaborate with others by considering the shared dynamics of the manipulated object using centralized training and decentralized execution (CTDE) multi-agent RL algorithms. When one agent interacts with the object, resulting in specific object dynamics changes, the other agents learn to respond appropriately, thereby achieving implicit communication and coordination between teammates. Unlike previous approaches that relied on tracking-based methods for multi-humanoid HOI, CooHOI is inherently efficient, does not depend on motion capture data of multi-humanoid interactions, and can be seamlessly extended to include more participants and a wide range of object types.

CooHOI: Learning Cooperative Human-Object Interaction with Manipulated Object Dynamics

[ "Jiawei Gao", "Ziqin Wang", "Zeqi Xiao", "Jingbo Wang", "Tai Wang", "Jinkun Cao", "Xiaolin Hu", "Si Liu", "Jifeng Dai", "Jiangmiao Pang" ]

NeurIPS.cc/2024/Conference

2406.14558

[ "" ]

-1

[]

0

oral

null

https://openreview.net/forum?id=iYkhThIXG1

@inproceedings{ cheng2024softlabel, title={Soft-Label Integration for Robust Toxicity Classification}, author={Zelei Cheng and Xian Wu and Jiahao Yu and Shuo Han and Xin-Qiang Cai and Xinyu Xing}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iYkhThIXG1} }

Toxicity classification in textual content remains a significant problem. Data with labels from a single annotator fall short of capturing the diversity of human perspectives. Therefore, there is a growing need to incorporate crowdsourced annotations for training an effective toxicity classifier. Additionally, the standard approach to training a classifier using empirical risk minimization (ERM) may fail to address the potential shifts between the training set and testing set due to exploiting spurious correlations. This work introduces a novel bi-level optimization framework that integrates crowdsourced annotations with the soft-labeling technique and optimizes the soft-label weights by Group Distributionally Robust Optimization (GroupDRO) to enhance the robustness against out-of-distribution (OOD) risk. We theoretically prove the convergence of our bi-level optimization algorithm. Experimental results demonstrate that our approach outperforms existing baseline methods in terms of both average and worst-group accuracy, confirming its effectiveness in leveraging crowdsourced annotations to achieve more effective and robust toxicity classification.

Soft-Label Integration for Robust Toxicity Classification

[ "Zelei Cheng", "Xian Wu", "Jiahao Yu", "Shuo Han", "Xin-Qiang Cai", "Xinyu Xing" ]

NeurIPS.cc/2024/Conference

2410.14894

[ "https://github.com/chengzelei/crowdsource_toxicity_classification" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iYcY7KAkSy

@inproceedings{ deng2024spiking, title={Spiking Token Mixer: A event-driven friendly Former structure for spiking neural networks}, author={Shikuang Deng and Yuhang Wu and Kangrui Du and Shi Gu}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iYcY7KAkSy} }

Spiking neural networks (SNNs), inspired by biological processes, use spike signals for inter-layer communication, presenting an energy-efficient alternative to traditional neural networks. To realize the theoretical advantages of SNNs in energy efficiency, it is essential to deploy them onto neuromorphic chips. On clock-driven synchronous chips, employing shorter time steps can enhance energy efficiency but reduce SNN performance. Compared to the clock-driven synchronous chip, the event-driven asynchronous chip achieves much lower energy consumption but only supports some specific network operations. Recently, a series of SNN projects have achieved tremendous success, significantly improving the SNN's performance. However, event-driven asynchronous chips do not support some of the proposed structures, making it impossible to integrate these SNNs into asynchronous hardware. In response to these problems, we propose the Spiking Token Mixer (STMixer) architecture, which consists exclusively of operations supported by asynchronous scenarios including convolutional, fully connected layers, and residual paths. Our series of experiments also demonstrate that STMixer achieves performance on par with spiking transformers in synchronous scenarios with very low timesteps. This indicates its ability to achieve the same level of performance with lower power consumption in synchronous scenarios. Codes are available at \url{https://github.com/brain-intelligence-lab/STMixer_demo}.

Spiking Token Mixer: A event-driven friendly Former structure for spiking neural networks

[ "Shikuang Deng", "Yuhang Wu", "Kangrui Du", "Shi Gu" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iWlqbNE8P7

@inproceedings{ huang2024physicsinformed, title={Physics-Informed Regularization for Domain-Agnostic Dynamical System Modeling}, author={Zijie Huang and Wanjia Zhao and Jingdong Gao and Ziniu Hu and Xiao Luo and Yadi Cao and Yuanzhou Chen and Yizhou Sun and Wei Wang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iWlqbNE8P7} }

Learning complex physical dynamics purely from data is challenging due to the intrinsic properties of systems to be satisfied. Incorporating physics-informed priors, such as in Hamiltonian Neural Networks (HNNs), achieves high-precision modeling for energy-conservative systems. However, real-world systems often deviate from strict energy conservation and follow different physical priors. To address this, we present a framework that achieves high-precision modeling for a wide range of dynamical systems from the numerical aspect, by enforcing Time-Reversal Symmetry (TRS) via a novel regularization term. It helps preserve energies for conservative systems while serving as a strong inductive bias for non-conservative, reversible systems. While TRS is a domain-specific physical prior, we present the first theoretical proof that TRS loss can universally improve modeling accuracy by minimizing higher-order Taylor terms in ODE integration, which is numerically beneficial to various systems regardless of their properties, even for irreversible systems. By integrating the TRS loss within neural ordinary differential equation models, the proposed model TREAT demonstrates superior performance on diverse physical systems. It achieves a significant 11.5% MSE improvement in a challenging chaotic triple-pendulum scenario, underscoring TREAT’s broad applicability and effectiveness.

Physics-Informed Regularization for Domain-Agnostic Dynamical System Modeling

[ "Zijie Huang", "Wanjia Zhao", "Jingdong Gao", "Ziniu Hu", "Xiao Luo", "Yadi Cao", "Yuanzhou Chen", "Yizhou Sun", "Wei Wang" ]

NeurIPS.cc/2024/Conference

2410.06366

[ "https://github.com/wanjiaZhao1203/TREAT" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iW0wXE0VyR

@inproceedings{ muneeb2024induced, title={Induced Model Matching: Restricted Models Help Train Full-Featured Models}, author={Usama Muneeb and Mesrob I Ohannessian}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iW0wXE0VyR} }

We consider scenarios where a very accurate (often small) predictive model using restricted features is available when training a full-featured (often larger) model. This restricted model may be thought of as ``side-information'', and can come either from an auxiliary dataset or from the same dataset by forcing the restriction. How can the restricted model be useful to the full model? To answer this, we introduce a methodology called Induced Model Matching (IMM). IMM aligns the context-restricted, or induced, version of the large model with the restricted model. We relate IMM to approaches such as noising, which is implicit in addressing the problem, and reverse knowledge distillation from weak teachers, which is explicit but does not exploit restriction being the nature of the weakness. We show that these prior methods can be thought of as approximations to IMM and can be problematic in terms of consistency. Experimentally, we first motivate IMM using logistic regression as a toy example. We then explore it in language modeling, the application that initially inspired it, and demonstrate it on both LSTM and transformer full models, using bigrams as restricted models. We lastly give a simple RL example, which shows that POMDP policies can help learn better MDP policies. The IMM principle is thus generally applicable in common scenarios where restricted data is cheaper to collect or restricted models are easier to learn.

Induced Model Matching: Restricted Models Help Train Full-Featured Models

[ "Usama Muneeb", "Mesrob I Ohannessian" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

oral

null

https://openreview.net/forum?id=iSjqTQ5S1f

@inproceedings{ vandenhirtz2024stochastic, title={Stochastic Concept Bottleneck Models}, author={Moritz Vandenhirtz and Sonia Laguna and Ri{\v{c}}ards Marcinkevi{\v{c}}s and Julia E Vogt}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iSjqTQ5S1f} }

Concept Bottleneck Models (CBMs) have emerged as a promising interpretable method whose final prediction is based on intermediate, human-understandable concepts rather than the raw input. Through time-consuming manual interventions, a user can correct wrongly predicted concept values to enhance the model's downstream performance. We propose *Stochastic Concept Bottleneck Models* (SCBMs), a novel approach that models concept dependencies. In SCBMs, a single-concept intervention affects all correlated concepts, thereby improving intervention effectiveness. Unlike previous approaches that model the concept relations via an autoregressive structure, we introduce an explicit, distributional parameterization that allows SCBMs to retain the CBMs' efficient training and inference procedure. Additionally, we leverage the parameterization to derive an effective intervention strategy based on the confidence region. We show empirically on synthetic tabular and natural image datasets that our approach improves intervention effectiveness significantly. Notably, we showcase the versatility and usability of SCBMs by examining a setting with CLIP-inferred concepts, alleviating the need for manual concept annotations.

Stochastic Concept Bottleneck Models

[ "Moritz Vandenhirtz", "Sonia Laguna", "Ričards Marcinkevičs", "Julia E Vogt" ]

NeurIPS.cc/2024/Conference

2406.19272

[ "https://github.com/mvandenhi/scbm" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iSfCWhvEGA

@inproceedings{ zheng2024learn, title={Learn To be Efficient: Build Structured Sparsity in Large Language Models}, author={Haizhong Zheng and Xiaoyan Bai and Xueshen Liu and Zhuoqing Mao and Beidi Chen and Fan Lai and Atul Prakash}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iSfCWhvEGA} }

Large Language Models (LLMs) have achieved remarkable success with their billion-level parameters, yet they incur high inference overheads. The emergence of activation sparsity in LLMs provides a natural approach to reduce this cost by involving only parts of the parameters for inference. However, existing methods only focus on utilizing this naturally formed activation sparsity in a post-training setting, overlooking the potential for further amplifying this inherent sparsity. In this paper, we hypothesize that LLMs can learn to be efficient by achieving more structured activation sparsity. To achieve this, we introduce a novel training algorithm, Learn-To-be-Efficient (LTE), designed to train efficiency-aware LLMs to learn to activate fewer neurons and achieve a better trade-off between sparsity and performance. Furthermore, unlike SOTA MoEfication methods, which mainly focus on ReLU-based models, LTE can also be applied to LLMs like LLaMA using non-ReLU activations. Extensive evaluation on language understanding, language generation, and instruction tuning tasks show that LTE consistently outperforms SOTA baselines. Along with our hardware-aware custom kernel implementation, LTE reduces LLaMA2-7B inference latency by 25% at 50% sparsity.

Learn To be Efficient: Build Structured Sparsity in Large Language Models

[ "Haizhong Zheng", "Xiaoyan Bai", "Xueshen Liu", "Zhuoqing Mao", "Beidi Chen", "Fan Lai", "Atul Prakash" ]

NeurIPS.cc/2024/Conference

2402.06126

[ "" ]

-1

[]

0

oral

null

https://openreview.net/forum?id=iSMTo0toDO

@inproceedings{ zhang2024subgdiff, title={SubgDiff: A Subgraph Diffusion Model to Improve Molecular Representation Learning}, author={Jiying Zhang and Zijing Liu and Yu Wang and Bin Feng and Yu Li}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iSMTo0toDO} }

Molecular representation learning has shown great success in advancing AI-based drug discovery. A key insight of many recent works is that the 3D geometric structure of molecules provides essential information about their physicochemical properties. Recently, denoising diffusion probabilistic models have achieved impressive performance in molecular 3D conformation generation. However, most existing molecular diffusion models treat each atom as an independent entity, overlooking the dependency among atoms within the substructures. This paper introduces a novel approach that enhances molecular representation learning by incorporating substructural information in the diffusion model framework. We propose a novel diffusion model termed SubgDiff for involving the molecular subgraph information in diffusion. Specifically, SubgDiff adopts three vital techniques: i) subgraph prediction, ii) expectation state, and iii) k-step same subgraph diffusion, to enhance the perception of molecular substructure in the denoising network. Experiments on extensive downstream tasks, especially the molecular force predictions, demonstrate the superior performance of our approach.

SubgDiff: A Subgraph Diffusion Model to Improve Molecular Representation Learning

[ "Jiying Zhang", "Zijing Liu", "Yu Wang", "Bin Feng", "Yu Li" ]

NeurIPS.cc/2024/Conference

2405.05665

[ "https://github.com/IDEA-XL/SubgDiff" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iQUxHrCna0

@inproceedings{ ypsilantis2024udon, title={{UDON}: Universal Dynamic Online distillatioN for generic image representations}, author={Nikolaos-Antonios Ypsilantis and Kaifeng Chen and Andre Araujo and Ondrej Chum}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iQUxHrCna0} }

Universal image representations are critical in enabling real-world fine-grained and instance-level recognition applications, where objects and entities from any domain must be identified at large scale. Despite recent advances, existing methods fail to capture important domain-specific knowledge, while also ignoring differences in data distribution across different domains. This leads to a large performance gap between efficient universal solutions and expensive approaches utilising a collection of specialist models, one for each domain. In this work, we make significant strides towards closing this gap, by introducing a new learning technique, dubbed UDON (Universal Dynamic Online distillatioN). UDON employs multi-teacher distillation, where each teacher is specialized in one domain, to transfer detailed domain-specific knowledge into the student universal embedding. UDON's distillation approach is not only effective, but also very efficient, by sharing most model parameters between the student and all teachers, where all models are jointly trained in an online manner. UDON also comprises a sampling technique which adapts the training process to dynamically allocate batches to domains which are learned slower and require more frequent processing. This boosts significantly the learning of complex domains which are characterised by a large number of classes and long-tail distributions. With comprehensive experiments, we validate each component of UDON, and showcase significant improvements over the state of the art in the recent UnED benchmark. Code: https://github.com/nikosips/UDON.

UDON: Universal Dynamic Online distillatioN for generic image representations

[ "Nikolaos-Antonios Ypsilantis", "Kaifeng Chen", "Andre Araujo", "Ondrej Chum" ]

NeurIPS.cc/2024/Conference

2406.08332

[ "https://github.com/nikosips/udon" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iOleSlC80F

@inproceedings{ li2024detmamba, title={3{DET}-Mamba: Causal Sequence Modelling for End-to-End 3D Object Detection}, author={Mingsheng Li and Jiakang Yuan and Sijin Chen and Lin Zhang and Anyu Zhu and Xin Chen and Tao Chen}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iOleSlC80F} }

Transformer-based architectures have been proven successful in detecting 3D objects from point clouds. However, the quadratic complexity of the attention mechanism struggles to encode rich information as point cloud resolution increases. Recently, state space models (SSM) such as Mamba have gained great attention due to their linear complexity and long sequence modeling ability for language understanding. To exploit the potential of Mamba on 3D scene-level perception, for the first time, we propose 3DET-Mamba, which is a novel SSM-based model designed for indoor 3d object detection. Specifically, we divide the point cloud into different patches and use a lightweight yet effective Inner Mamba to capture local geometric information. To observe the scene from a global perspective, we introduce a novel Dual Mamba module that models the point cloud in terms of spatial distribution and continuity. Additionally, we design a Query-aware Mamba module that decodes context features into object sets under the guidance of learnable queries. Extensive experiments demonstrate that 3DET-Mamba surpasses previous 3DETR on indoor 3D detection benchmarks such as ScanNet, improving AP25/AP50 from 65.0\%/47.0\% to 70.4\%/54.4\%, respectively.

3DET-Mamba: Causal Sequence Modelling for End-to-End 3D Object Detection

[ "Mingsheng Li", "Jiakang Yuan", "Sijin Chen", "Lin Zhang", "Anyu Zhu", "Xin Chen", "Tao Chen" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iO7viYaAt7

@inproceedings{ mechergui2024expectation, title={Expectation Alignment: Handling Reward Misspecification in the Presence of Expectation Mismatch}, author={Malek Mechergui and Sarath Sreedharan}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iO7viYaAt7} }

Detecting and handling misspecified objectives, such as reward functions, has been widely recognized as one of the central challenges within the domain of Artificial Intelligence (AI) safety research. However, even with the recognition of the importance of this problem, we are unaware of any works that attempt to provide a clear definition for what constitutes (a) misspecified objectives and (b) successfully resolving such misspecifications. In this work, we use the theory of mind, i.e., the human user's beliefs about the AI agent, as a basis to develop a formal explanatory framework, called Expectation Alignment (EAL), to understand the objective misspecification and its causes. Our EAL framework not only acts as an explanatory framework for existing works but also provides us with concrete insights into the limitations of existing methods to handle reward misspecification and novel solution strategies. We use these insights to propose a new interactive algorithm that uses the specified reward to infer potential user expectations about the system behavior. We show how one can efficiently implement this algorithm by mapping the inference problem into linear programs. We evaluate our method on a set of standard Markov Decision Process (MDP) benchmarks.

Expectation Alignment: Handling Reward Misspecification in the Presence of Expectation Mismatch

[ "Malek Mechergui", "Sarath Sreedharan" ]

NeurIPS.cc/2024/Conference

2404.08791

[ "https://github.com/malek-mechergui/codemdp" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iO6tcLJEwA

@inproceedings{ min2024epipolarfree, title={Epipolar-Free 3D Gaussian Splatting for Generalizable Novel View Synthesis}, author={Zhiyuan Min and Yawei Luo and Jianwen Sun and Yi Yang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iO6tcLJEwA} }

Generalizable 3D Gaussian splitting (3DGS) can reconstruct new scenes from sparse-view observations in a feed-forward inference manner, eliminating the need for scene-specific retraining required in conventional 3DGS. However, existing methods rely heavily on epipolar priors, which can be unreliable in complex real-world scenes, particularly in non-overlapping and occluded regions. In this paper, we propose eFreeSplat, an efficient feed-forward 3DGS-based model for generalizable novel view synthesis that operates independently of epipolar line constraints. To enhance multiview feature extraction with 3D perception, we employ a self-supervised Vision Transformer (ViT) with cross-view completion pre-training on large-scale datasets. Additionally, we introduce an Iterative Cross-view Gaussians Alignment method to ensure consistent depth scales across different views. Our eFreeSplat represents a new paradigm for generalizable novel view synthesis. We evaluate eFreeSplat on wide-baseline novel view synthesis tasks using the RealEstate10K and ACID datasets. Extensive experiments demonstrate that eFreeSplat surpasses state-of-the-art baselines that rely on epipolar priors, achieving superior geometry reconstruction and novel view synthesis quality.

Epipolar-Free 3D Gaussian Splatting for Generalizable Novel View Synthesis

[ "Zhiyuan Min", "Yawei Luo", "Jianwen Sun", "Yi Yang" ]

NeurIPS.cc/2024/Conference

2410.22817

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iNvXYQrkpi

@inproceedings{ zhan2024fast, title={Fast and Memory-Efficient Video Diffusion Using Streamlined Inference}, author={Zheng Zhan and Yushu Wu and Yifan Gong and Zichong Meng and Zhenglun Kong and Changdi Yang and Geng Yuan and Pu Zhao and Wei Niu and Yanzhi Wang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iNvXYQrkpi} }

The rapid progress in artificial intelligence-generated content (AIGC), especially with diffusion models, has significantly advanced development of high-quality video generation. However, current video diffusion models exhibit demanding computational requirements and high peak memory usage, especially for generating longer and higher-resolution videos. These limitations greatly hinder the practical application of video diffusion models on standard hardware platforms. To tackle this issue, we present a novel, training-free framework named Streamlined Inference, which leverages the temporal and spatial properties of video diffusion models. Our approach integrates three core components: Feature Slicer, Operator Grouping, and Step Rehash. Specifically, Feature Slicer effectively partitions input features into sub-features and Operator Grouping processes each sub-feature with a group of consecutive operators, resulting in significant memory reduction without sacrificing the quality or speed. Step Rehash further exploits the similarity between adjacent steps in diffusion, and accelerates inference through skipping unnecessary steps. Extensive experiments demonstrate that our approach significantly reduces peak memory and computational overhead, making it feasible to generate high-quality videos on a single consumer GPU (e.g., reducing peak memory of Animatediff from 42GB to 11GB, featuring faster inference on 2080Ti).

Fast and Memory-Efficient Video Diffusion Using Streamlined Inference

[ "Zheng Zhan", "Yushu Wu", "Yifan Gong", "Zichong Meng", "Zhenglun Kong", "Changdi Yang", "Geng Yuan", "Pu Zhao", "Wei Niu", "Yanzhi Wang" ]

NeurIPS.cc/2024/Conference

2411.01171

[ "https://github.com/wuyushuwys/FMEDiffusion" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iNUKoLU8xb

@inproceedings{ chen2024your, title={Your contrastive learning problem is secretly a distribution alignment problem}, author={Zihao Chen and Chi-Heng Lin and Ran Liu and Jingyun Xiao and Eva L Dyer}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iNUKoLU8xb} }

Despite the success of contrastive learning (CL) in vision and language, its theoretical foundations and mechanisms for building representations remain poorly understood. In this work, we build connections between noise contrastive estimation losses widely used in CL and distribution alignment with entropic optimal transport (OT). This connection allows us to develop a family of different losses and multistep iterative variants for existing CL methods. Intuitively, by using more information from the distribution of latents, our approach allows a more distribution-aware manipulation of the relationships within augmented sample sets. We provide theoretical insights and experimental evidence demonstrating the benefits of our approach for generalized contrastive alignment. Through this framework, it is possible to leverage tools in OT to build unbalanced losses to handle noisy views and customize the representation space by changing the constraints on alignment. By reframing contrastive learning as an alignment problem and leveraging existing optimization tools for OT, our work provides new insights and connections between different self-supervised learning models in addition to new tools that can be more easily adapted to incorporate domain knowledge into learning.

Your contrastive learning problem is secretly a distribution alignment problem

[ "Zihao Chen", "Chi-Heng Lin", "Ran Liu", "Jingyun Xiao", "Eva L Dyer" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iNS3SC949v

@inproceedings{ castro-mac{\'\i}as2024sm, title={Sm: enhanced localization in Multiple Instance Learning for medical imaging classification}, author={Francisco M Castro-Mac{\'\i}as and Pablo Morales-Alvarez and Yunan Wu and Rafael Molina and Aggelos Katsaggelos}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iNS3SC949v} }

Multiple Instance Learning (MIL) is widely used in medical imaging classification to reduce the labeling effort. While only bag labels are available for training, one typically seeks predictions at both bag and instance levels (classification and localization tasks, respectively). Early MIL methods treated the instances in a bag independently. Recent methods account for global and local dependencies among instances. Although they have yielded excellent results in classification, their performance in terms of localization is comparatively limited. We argue that these models have been designed to target the classification task, while implications at the instance level have not been deeply investigated. Motivated by a simple observation -- that neighboring instances are likely to have the same label -- we propose a novel, principled, and flexible mechanism to model local dependencies. It can be used alone or combined with any mechanism to model global dependencies (e.g., transformers). A thorough empirical validation shows that our module leads to state-of-the-art performance in localization while being competitive or superior in classification. Our code is at https://github.com/Franblueee/SmMIL.

Sm: enhanced localization in Multiple Instance Learning for medical imaging classification

[ "Francisco M Castro-Macías", "Pablo Morales-Alvarez", "Yunan Wu", "Rafael Molina", "Aggelos Katsaggelos" ]

NeurIPS.cc/2024/Conference

2410.03276

[ "https://github.com/franblueee/smmil" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iN43sJoib7

@inproceedings{ kim2024are, title={Are Self-Attentions Effective for Time Series Forecasting?}, author={Dongbin Kim and Jinseong Park and Jaewook Lee and Hoki Kim}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iN43sJoib7} }

Time series forecasting is crucial for applications across multiple domains and various scenarios. Although Transformers have dramatically advanced the landscape of forecasting, their effectiveness remains debated. Recent findings have indicated that simpler linear models might outperform complex Transformer-based approaches, highlighting the potential for more streamlined architectures. In this paper, we shift the focus from evaluating the overall Transformer architecture to specifically examining the effectiveness of self-attention for time series forecasting. To this end, we introduce a new architecture, Cross-Attention-only Time Series transformer (CATS), that rethinks the traditional transformer framework by eliminating self-attention and leveraging cross-attention mechanisms instead. By establishing future horizon-dependent parameters as queries and enhanced parameter sharing, our model not only improves long-term forecasting accuracy but also reduces the number of parameters and memory usage. Extensive experiment across various datasets demonstrates that our model achieves superior performance with the lowest mean squared error and uses fewer parameters compared to existing models. The implementation of our model is available at: https://github.com/dongbeank/CATS.

Are Self-Attentions Effective for Time Series Forecasting?

[ "Dongbin Kim", "Jinseong Park", "Jaewook Lee", "Hoki Kim" ]

NeurIPS.cc/2024/Conference

2405.16877

[ "https://github.com/dongbeank/cats" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iMEAHXDiNP

@inproceedings{ tullii2024improved, title={Improved Algorithms for Contextual Dynamic Pricing}, author={Matilde Tullii and Solenne Gaucher and Nadav Merlis and Vianney Perchet}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iMEAHXDiNP} }

In contextual dynamic pricing, a seller sequentially prices goods based on contextual information. Buyers will purchase products only if the prices are below their valuations. The goal of the seller is to design a pricing strategy that collects as much revenue as possible. We focus on two different valuation models. The first assumes that valuations linearly depend on the context and are further distorted by noise. Under minor regularity assumptions, our algorithm achieves an optimal regret bound of $\tilde{\mathcal{O}}(T^{2/3})$, improving the existing results. The second model removes the linearity assumption, requiring only that the expected buyer valuation is $\beta$-H\"older in the context. For this model, our algorithm obtains a regret $\tilde{\mathcal{O}}(T^{d+2\beta/d+3\beta})$, where $d$ is the dimension of the context space.

Improved Algorithms for Contextual Dynamic Pricing

[ "Matilde Tullii", "Solenne Gaucher", "Nadav Merlis", "Vianney Perchet" ]

NeurIPS.cc/2024/Conference

2406.11316

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iM5i289eqt

@inproceedings{ qian2024maskfactory, title={MaskFactory: Towards High-quality Synthetic Data Generation for Dichotomous Image Segmentation}, author={Haotian Qian and Yinda Chen and Shengtao Lou and Fahad Khan and Xiaogang Jin and Deng-Ping Fan}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iM5i289eqt} }

Dichotomous Image Segmentation (DIS) tasks require highly precise annotations, and traditional dataset creation methods are labor intensive, costly, and require extensive domain expertise. Although using synthetic data for DIS is a promising solution to these challenges, current generative models and techniques struggle with the issues of scene deviations, noise-induced errors, and limited training sample variability. To address these issues, we introduce a novel approach, Mask Factory, which provides a scalable solution for generating diverse and precise datasets, markedly reducing preparation time and costs. We first introduce a general mask editing method that combines rigid and non-rigid editing techniques to generate high-quality synthetic masks. Specially, rigid editing leverages geometric priors from diffusion models to achieve precise viewpoint transformations under zero-shot conditions, while non-rigid editing employs adversarial training and self-attention mechanisms for complex, topologically consistent modifications. Then, we generate pairs of high-resolution image and accurate segmentation mask using a multi-conditional control generation method. Finally, our experiments on the widely-used DIS5K dataset benchmark demonstrate superior performance in quality and efficiency compared to existing methods. The code is available at https://qian-hao-tian.github.io/MaskFactory/.

MaskFactory: Towards High-quality Synthetic Data Generation for Dichotomous Image Segmentation

[ "Haotian Qian", "Yinda Chen", "Shengtao Lou", "Fahad Khan", "Xiaogang Jin", "Deng-Ping Fan" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iJgwd5mWYg

@inproceedings{ bernasconi2024beyond, title={Beyond Primal-Dual Methods in Bandits with Stochastic and Adversarial Constraints}, author={Martino Bernasconi and Matteo Castiglioni and Andrea Celli and Federico Fusco}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iJgwd5mWYg} }

We address a generalization of the bandit with knapsacks problem, where a learner aims to maximize rewards while satisfying an arbitrary set of long-term constraints. Our goal is to design best-of-both-worlds algorithms that perform optimally under both stochastic and adversarial constraints. Previous works address this problem via primal-dual methods, and require some stringent assumptions, namely the Slater's condition, and in adversarial settings, they either assume knowledge of a lower bound on the Slater's parameter, or impose strong requirements on the primal and dual regret minimizers such as requiring weak adaptivity. We propose an alternative and more natural approach based on optimistic estimations of the constraints. Surprisingly, we show that estimating the constraints with an UCB-like approach guarantees optimal performances. Our algorithm consists of two main components: (i) a regret minimizer working on moving strategy sets and (ii) an estimate of the feasible set as an optimistic weighted empirical mean of previous samples. The key challenge in this approach is designing adaptive weights that meet the different requirements for stochastic and adversarial constraints. Our algorithm is significantly simpler than previous approaches, and has a cleaner analysis. Moreover, ours is the first best-of-both-worlds algorithm providing bounds logarithmic in the number of constraints. Additionally, in stochastic settings, it provides $\widetilde O(\sqrt{T})$ regret without Slater's condition.

Beyond Primal-Dual Methods in Bandits with Stochastic and Adversarial Constraints

[ "Martino Bernasconi", "Matteo Castiglioni", "Andrea Celli", "Federico Fusco" ]

NeurIPS.cc/2024/Conference

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iFKmFUxQDh

@inproceedings{ guo2024refir, title={Re{FIR}: Grounding Large Restoration Models with Retrieval Augmentation}, author={Hang Guo and Tao Dai and Zhihao Ouyang and Taolin Zhang and Yaohua Zha and Bin Chen and Shu-Tao Xia}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iFKmFUxQDh} }

Recent advances in diffusion-based Large Restoration Models (LRMs) have significantly improved photo-realistic image restoration by leveraging the internal knowledge embedded within model weights. However, existing LRMs often suffer from the hallucination dilemma, i.e., producing incorrect contents or textures when dealing with severe degradations, due to their heavy reliance on limited internal knowledge. In this paper, we propose an orthogonal solution called the Retrieval-augmented Framework for Image Restoration (ReFIR), which incorporates retrieved images as external knowledge to extend the knowledge boundary of existing LRMs in generating details faithful to the original scene. Specifically, we first introduce the nearest neighbor lookup to retrieve content-relevant high-quality images as reference, after which we propose the cross-image injection to modify existing LRMs to utilize high-quality textures from retrieved images. Thanks to the additional external knowledge, our ReFIR can well handle the hallucination challenge and facilitate faithfully results. Extensive experiments demonstrate that ReFIR can achieve not only high-fidelity but also realistic restoration results. Importantly, our ReFIR requires no training and is adaptable to various LRMs.

ReFIR: Grounding Large Restoration Models with Retrieval Augmentation

[ "Hang Guo", "Tao Dai", "Zhihao Ouyang", "Taolin Zhang", "Yaohua Zha", "Bin Chen", "Shu-Tao Xia" ]

NeurIPS.cc/2024/Conference

2410.05601

[ "https://github.com/csguoh/refir" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iF7MnXnxRw

@inproceedings{ sieber2024understanding, title={Understanding the Differences in Foundation Models: Attention, State Space Models, and Recurrent Neural Networks}, author={Jerome Sieber and Carmen Amo Alonso and Alexandre Didier and Melanie Zeilinger and Antonio Orvieto}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iF7MnXnxRw} }

Softmax attention is the principle backbone of foundation models for various artificial intelligence applications, yet its quadratic complexity in sequence length can limit its inference throughput in long-context settings. To address this challenge, alternative architectures such as linear attention, State Space Models (SSMs), and Recurrent Neural Networks (RNNs) have been considered as more efficient alternatives. While connections between these approaches exist, such models are commonly developed in isolation and there is a lack of theoretical understanding of the shared principles underpinning these architectures and their subtle differences, greatly influencing performance and scalability. In this paper, we introduce the Dynamical Systems Framework (DSF), which allows a principled investigation of all these architectures in a common representation. Our framework facilitates rigorous comparisons, providing new insights on the distinctive characteristics of each model class. For instance, we compare linear attention and selective SSMs, detailing their differences and conditions under which both are equivalent. We also provide principled comparisons between softmax attention and other model classes, discussing the theoretical conditions under which softmax attention can be approximated. Additionally, we substantiate these new insights with empirical validations and mathematical arguments. This shows the DSF's potential to guide the systematic development of future more efficient and scalable foundation models.

Understanding the Differences in Foundation Models: Attention, State Space Models, and Recurrent Neural Networks

[ "Jerome Sieber", "Carmen Amo Alonso", "Alexandre Didier", "Melanie Zeilinger", "Antonio Orvieto" ]

NeurIPS.cc/2024/Conference

[ "https://github.com/intelligentcontrolsystems/dsf-mqar" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iEsyRsg6t1

@inproceedings{ abouei2024causal, title={Causal Effect Identification in a Sub-Population with Latent Variables}, author={Amir Mohammad Abouei and Ehsan Mokhtarian and Negar Kiyavash and Matthias Grossglauser}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iEsyRsg6t1} }

The s-ID problem seeks to compute a causal effect in a specific sub-population from the observational data pertaining to the same sub population (Abouei et al., 2023). This problem has been addressed when all the variables in the system are observable. In this paper, we consider an extension of the s-ID problem that allows for the presence of latent variables. To tackle the challenges induced by the presence of latent variables in a sub-population, we first extend the classical relevant graphical definitions, such as c-components and Hedges, initially defined for the so-called ID problem (Pearl, 1995; Tian & Pearl, 2002), to their new counterparts. Subsequently, we propose a sound algorithm for the s-ID problem with latent variables.

Causal Effect Identification in a Sub-Population with Latent Variables

[ "Amir Mohammad Abouei", "Ehsan Mokhtarian", "Negar Kiyavash", "Matthias Grossglauser" ]

NeurIPS.cc/2024/Conference

2405.14547

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iEeiZlTbts

@inproceedings{ rutherford2024no, title={No Regrets: Investigating and Improving Regret Approximations for Curriculum Discovery}, author={Alexander Rutherford and Michael Beukman and Timon Willi and Bruno Lacerda and Nick Hawes and Jakob Nicolaus Foerster}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iEeiZlTbts} }

What data or environments to use for training to improve downstream performance is a longstanding and very topical question in reinforcement learning. In particular, Unsupervised Environment Design (UED) methods have gained recent attention as their adaptive curricula promise to enable agents to be robust to in- and out-of-distribution tasks. This work investigates how existing UED methods select training environments, focusing on task prioritisation metrics. Surprisingly, despite methods aiming to maximise regret in theory, the practical approximations do not correlate with regret but with success rate. As a result, a significant portion of an agent's experience comes from environments it has already mastered, offering little to no contribution toward enhancing its abilities. Put differently, current methods fail to predict intuitive measures of *learnability*. Specifically, they are unable to consistently identify those scenarios that the agent can sometimes solve, but not always. Based on our analysis, we develop a method that directly trains on scenarios with high learnability. This simple and intuitive approach outperforms existing UED methods in several binary-outcome environments, including the standard domain of Minigrid and a novel setting closely inspired by a real-world robotics problem. We further introduce a new adversarial evaluation procedure for directly measuring robustness, closely mirroring the conditional value at risk (CVaR). We open-source all our code and present visualisations of final policies here: https://github.com/amacrutherford/sampling-for-learnability.

No Regrets: Investigating and Improving Regret Approximations for Curriculum Discovery

[ "Alexander Rutherford", "Michael Beukman", "Timon Willi", "Bruno Lacerda", "Nick Hawes", "Jakob Nicolaus Foerster" ]

NeurIPS.cc/2024/Conference

2408.15099

[ "https://github.com/amacrutherford/sampling-for-learnability" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iD18l6prA7

@inproceedings{ crisostomi2024cm, title={\$C{\textasciicircum}2M{\textasciicircum}3\$: Cycle-Consistent Multi-Model Merging}, author={Donato Crisostomi and Marco Fumero and Daniele Baieri and Florian Bernard and Emanuele Rodol{\`a}}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iD18l6prA7} }

In this paper, we present a novel data-free method for merging neural networks in weight space. Our method optimizes for the permutations of network neurons while ensuring global coherence across all layers, and it outperforms recent layer-local approaches in a set of challenging scenarios. We then generalize the formulation to the $N$-models scenario to enforce cycle consistency of the permutations with guarantees, allowing circular compositions of permutations to be computed without accumulating error along the path. We qualitatively and quantitatively motivate the need for such a constraint, showing its benefits when merging homogeneous sets of models in scenarios spanning varying architectures and datasets. We finally show that, when coupled with activation renormalization, the approach yields the best results in the task.

C^2M^3: Cycle-Consistent Multi-Model Merging

[ "Donato Crisostomi", "Marco Fumero", "Daniele Baieri", "Florian Bernard", "Emanuele Rodolà" ]

NeurIPS.cc/2024/Conference

[ "https://github.com/crisostomi/cycle-consistent-model-merging" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iChQIJtjHB

@inproceedings{ zhu2024ssos, title={S-{SOS}: Stochastic Sum-Of-Squares for Parametric Polynomial Optimization}, author={Richard Licheng Zhu and Mathias Oster and Yuehaw Khoo}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iChQIJtjHB} }

Global polynomial optimization is an important tool across applied mathematics, with many applications in operations research, engineering, and the physical sciences. In various settings, the polynomials depend on external parameters that may be random. We discuss a stochastic sum-of-squares (S-SOS) algorithm based on the sum-of-squares hierarchy that constructs a series of semidefinite programs to jointly find strict lower bounds on the global minimum and extracts candidates for parameterized global minimizers. We prove quantitative convergence of the hierarchy as the degree increases and use it to solve unconstrained and constrained polynomial optimization problems parameterized by random variables. By employing n-body priors from condensed matter physics to induce sparsity, we can use S-SOS to produce solutions and uncertainty intervals for sensor network localization problems containing up to 40 variables and semidefinite matrix sizes surpassing 800 x 800.

S-SOS: Stochastic Sum-Of-Squares for Parametric Polynomial Optimization

[ "Richard Licheng Zhu", "Mathias Oster", "Yuehaw Khoo" ]

NeurIPS.cc/2024/Conference

2406.08954

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iC869BBmc5

@inproceedings{ chen2024proedit, title={ProEdit: Simple Progression is All You Need for High-Quality 3D Scene Editing}, author={Jun-Kun Chen and Yu-Xiong Wang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iC869BBmc5} }

This paper proposes ProEdit - a simple yet effective framework for high-quality 3D scene editing guided by diffusion distillation in a novel progressive manner. Inspired by the crucial observation that multi-view inconsistency in scene editing is rooted in the diffusion model’s large feasible output space (FOS), our framework controls the size of FOS and reduces inconsistency by decomposing the overall editing task into several subtasks, which are then executed progressively on the scene. Within this framework, we design a difficulty-aware subtask decomposition scheduler and an adaptive 3D Gaussian splatting (3DGS) training strategy, ensuring high efficiency in performing each subtask. Extensive evaluation shows that our ProEdit achieves state-of-the-art results in various scenes and challenging editing tasks, all through a simple framework without any expensive or sophisticated add-ons like distillation losses, components, or training procedures. Notably, ProEdit also provides a new way to preview, control, and select the aggressivity of editing operation during the editing process.

ProEdit: Simple Progression is All You Need for High-Quality 3D Scene Editing

[ "Jun-Kun Chen", "Yu-Xiong Wang" ]

NeurIPS.cc/2024/Conference

2411.05006

[ "" ]

-1

[]

0

poster

null

https://openreview.net/forum?id=iBZSOh027z

@inproceedings{ song2024similaritynavigated, title={Similarity-Navigated Conformal Prediction for Graph Neural Networks}, author={Jianqing Song and Jianguo Huang and Wenyu Jiang and Baoming Zhang and Shuangjie Li and Chongjun Wang}, booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems}, year={2024}, url={https://openreview.net/forum?id=iBZSOh027z} }

Graph Neural Networks have achieved remarkable accuracy in semi-supervised node classification tasks. However, these results lack reliable uncertainty estimates. Conformal prediction methods provide a theoretical guarantee for node classification tasks, ensuring that the conformal prediction set contains the ground-truth label with a desired probability (e.g., 95\%). In this paper, we empirically show that for each node, aggregating the non-conformity scores of nodes with the same label can improve the efficiency of conformal prediction sets while maintaining valid marginal coverage. This observation motivates us to propose a novel algorithm named $\textit{Similarity-Navigated Adaptive Prediction Sets}$ (SNAPS), which aggregates the non-conformity scores based on feature similarity and structural neighborhood. The key idea behind SNAPS is that nodes with high feature similarity or direct connections tend to have the same label. By incorporating adaptive similar nodes information, SNAPS can generate compact prediction sets and increase the singleton hit ratio (correct prediction sets of size one). Moreover, we theoretically provide a finite-sample coverage guarantee of SNAPS. Extensive experiments demonstrate the superiority of SNAPS, improving the efficiency of prediction sets and singleton hit ratio while maintaining valid coverage.

Similarity-Navigated Conformal Prediction for Graph Neural Networks

[ "Jianqing Song", "Jianguo Huang", "Wenyu Jiang", "Baoming Zhang", "Shuangjie Li", "Chongjun Wang" ]

NeurIPS.cc/2024/Conference

2405.14303

[ "" ]

-1

[]

0

poster