id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2312.03919 | Kenneth Gill | Kenneth Gill | Indivisibility and uniform computational strength | null | null | null | null | math.LO cs.LO math.CO | http://creativecommons.org/licenses/by/4.0/ | A countable structure is indivisible if for every coloring with finite range there is a monochromatic isomorphic subcopy of the structure. Each indivisible structure naturally corresponds to an indivisibility problem which outputs such a subcopy given a presentation and coloring. We investigate the Weihrauch complexity of the indivisibility problems for two structures: the rational numbers $\mathbb{Q}$ as a linear order, and the equivalence relation $\mathscr{E}$ with countably many equivalence classes each having countably many members. We separate the Weihrauch degrees of both corresponding indivisibility problems from several benchmarks, showing in particular that the indivisibility problem for $\mathbb{Q}$ cannot solve the problem of finding a monochromatic rational interval given a coloring for which there is one; and that the Weihrauch degree of the indivisibility problem for $\mathscr{E}$ is strictly between those of $\mathsf{RT}^2$ and $\mathsf{SRT}^2$, two widely studied variants of Ramsey's theorem for pairs whose reverse-mathematical separation was open until recently. | [
{
"version": "v1",
"created": "Wed, 06 Dec 2023 21:34:37 GMT"
},
{
"version": "v2",
"created": "Fri, 08 Dec 2023 16:30:35 GMT"
},
{
"version": "v3",
"created": "Wed, 13 Dec 2023 00:55:46 GMT"
},
{
"version": "v4",
"created": "Thu, 11 Jan 2024 18:36:47 GMT"
},
{
"version": "v5",
"created": "Mon, 06 May 2024 23:48:41 GMT"
},
{
"version": "v6",
"created": "Tue, 29 Oct 2024 12:58:05 GMT"
},
{
"version": "v7",
"created": "Sun, 16 Mar 2025 22:48:18 GMT"
},
{
"version": "v8",
"created": "Sat, 07 Jun 2025 08:14:39 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Gill",
"Kenneth",
""
]
] |
2312.03941 | Hritika Gupta | Mark Fackrell, Hritika Gupta, Peter G. Taylor | Minimising Numbers of Losses and Abandonments in Small Call Centres Under a Transient Regime | null | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | We consider the problem of allocating customers to agents in small call centres so that transient performance indicators in terms of expected numbers of losses and abandonments are optimised. To gain insight into the general structure of optimal policies, we start by (i) using backward induction based upon Bellman's equation for a finite-horizon discrete-time model, and (ii) deriving stationary policies for an infinite-horizon continuous-time model with discounting. We then compare the performance of such policies in our original finite-horizon continuous-time model using a first-step analysis applied to the Laplace transforms of the relevant measures. | [
{
"version": "v1",
"created": "Wed, 06 Dec 2023 22:47:05 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 01:46:31 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Fackrell",
"Mark",
""
],
[
"Gupta",
"Hritika",
""
],
[
"Taylor",
"Peter G.",
""
]
] |
2312.04398 | Yongqi Dong | Yongqi Dong, Xingmin Lu, Ruohan Li, Wei Song, Bart van Arem, Haneen Farah | Intelligent Anomaly Detection for Lane Rendering Using Transformer with Self-Supervised Pre-Training and Customized Fine-Tuning | 16 pages, 7 figures, accepted and presented at the 103rd Transportation Research Board (TRB) Annual Meeting, and published by Transportation Research Record: Journal of the Transportation Research Board | null | 10.1177/03611981251333341 | null | cs.CV cs.AI cs.LG eess.IV stat.ML | http://creativecommons.org/licenses/by/4.0/ | The burgeoning navigation services using digital maps provide great convenience to drivers. Nevertheless, the presence of anomalies in lane rendering map images occasionally introduces potential hazards, as such anomalies can be misleading to human drivers and consequently contribute to unsafe driving conditions. In response to this concern and to accurately and effectively detect the anomalies, this paper transforms lane rendering image anomaly detection into a classification problem and proposes a four-phase pipeline consisting of data pre-processing, self-supervised pre-training with the masked image modeling (MiM) method, customized fine-tuning using cross-entropy based loss with label smoothing, and post-processing to tackle it leveraging state-of-the-art deep learning techniques, especially those involving Transformer models. Various experiments verify the effectiveness of the proposed pipeline. Results indicate that the proposed pipeline exhibits superior performance in lane rendering image anomaly detection, and notably, the self-supervised pre-training with MiM can greatly enhance the detection accuracy while significantly reducing the total training time. For instance, employing the Swin Transformer with Uniform Masking as self-supervised pretraining (Swin-Trans-UM) yielded a heightened accuracy at 94.77% and an improved Area Under The Curve (AUC) score of 0.9743 compared with the pure Swin Transformer without pre-training (Swin-Trans) with an accuracy of 94.01% and an AUC of 0.9498. The fine-tuning epochs were dramatically reduced to 41 from the original 280. In conclusion, the proposed pipeline, with its incorporation of self-supervised pre-training using MiM and other advanced deep learning techniques, emerges as a robust solution for enhancing the accuracy and efficiency of lane rendering image anomaly detection in digital navigation systems. | [
{
"version": "v1",
"created": "Thu, 07 Dec 2023 16:10:10 GMT"
},
{
"version": "v2",
"created": "Wed, 29 May 2024 15:54:04 GMT"
},
{
"version": "v3",
"created": "Mon, 25 Nov 2024 12:26:11 GMT"
},
{
"version": "v4",
"created": "Wed, 27 Nov 2024 00:21:57 GMT"
},
{
"version": "v5",
"created": "Thu, 20 Feb 2025 05:58:40 GMT"
},
{
"version": "v6",
"created": "Sat, 07 Jun 2025 22:29:37 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Dong",
"Yongqi",
""
],
[
"Lu",
"Xingmin",
""
],
[
"Li",
"Ruohan",
""
],
[
"Song",
"Wei",
""
],
[
"van Arem",
"Bart",
""
],
[
"Farah",
"Haneen",
""
]
] |
2312.09272 | Christopher Fechisin | Christopher Fechisin, Nathanan Tantivasadakarn, Victor V. Albert | Non-invertible symmetry-protected topological order in a group-based cluster state | 30+13 pages, 7 figures. v2: added Sec IV.C, Appendix I, and several references | Phys. Rev. X 15, 011058 (2025) | 10.1103/PhysRevX.15.011058 | null | cond-mat.str-el hep-th quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite growing interest in beyond-group symmetries in quantum condensed matter systems, there are relatively few microscopic lattice models explicitly realizing these symmetries, and many phenomena have yet to be studied at the microscopic level. We introduce a one-dimensional stabilizer Hamiltonian composed of group-based Pauli operators whose ground state is a $G\times \text{Rep}(G)$-symmetric state: the $G \textit{ cluster state}$ introduced in Brell, New Journal of Physics 17, 023029 (2015) [at http://doi.org/10.1088/1367-2630/17/2/023029]. We show that this state lies in a symmetry-protected topological (SPT) phase protected by $G\times \text{Rep}(G)$ symmetry, distinct from the symmetric product state by a duality argument. We identify several signatures of SPT order, namely protected edge modes, string order parameters, and topological response. We discuss how $G$ cluster states may be used as a universal resource for measurement-based quantum computation, explicitly working out the case where $G$ is a semidirect product of abelian groups. | [
{
"version": "v1",
"created": "Thu, 14 Dec 2023 18:56:20 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Jun 2024 06:35:31 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Fechisin",
"Christopher",
""
],
[
"Tantivasadakarn",
"Nathanan",
""
],
[
"Albert",
"Victor V.",
""
]
] |
2312.10290 | Weijie Zheng | Weijie Zheng, Benjamin Doerr | Runtime Analysis of the SMS-EMOA for Many-Objective Optimization | Improved full version of a paper accepted in AAAI 2024 | Conference on Artificial Intelligence, AAAI 2024, pages 20874-20882 | 10.1609/AAAI.V38I18.30077 | null | cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper conducts the first rigorous runtime analysis of the SMS-EMOA for many-objective optimization. To this aim, we first propose a many-objective counterpart of the bi-objective OJZJ benchmark. We prove that SMS-EMOA computes the full Pareto front of this benchmark in an expected number of $O(\mu M n^k)$ iterations, where $n$ denotes the problem size (length of the bit-string representation), $k$ the gap size (a difficulty parameter of the problem), $M=(2n/m-2k+3)^{m/2}$ the size of the Pareto front, and $\mu$ the population size (at least the same size as the largest incomparable set). This result together with the existing negative result for the original NSGA-II shows that, in principle, the general approach of the NSGA-II is suitable for many-objective optimization, but the crowding distance as tie-breaker has deficiencies.
We obtain three additional insights on the SMS-EMOA. Different from a recent result for the bi-objective \ojzj benchmark, a recently proposed stochastic population update often does not help for its many-objective counterpart. It at most results in a speed-up by a factor of order $2^{k} / \mu$, which is $\Theta(1)$ for large $m$, such as $m>k$. On the positive side, we prove that heavy-tailed mutation irrespective of the number $m$ of objectives results in a speed-up of order $k^{0.5+k-\beta}/e^k$. Finally, we conduct the first runtime analyses of the SMS-EMOA on the classic OMM and LOTZ and show that the SMS-EMOA has a performance comparable to the GSEMO and the NSGA-II.
Our main technical insight, a general condition ensuring that the SMS-EMOA does not lose Pareto-optimal objective values, promises to be useful also in other runtime analyses of this algorithm. | [
{
"version": "v1",
"created": "Sat, 16 Dec 2023 02:23:09 GMT"
},
{
"version": "v2",
"created": "Thu, 25 Jan 2024 09:47:42 GMT"
},
{
"version": "v3",
"created": "Wed, 12 Jun 2024 12:55:13 GMT"
},
{
"version": "v4",
"created": "Sat, 07 Jun 2025 03:04:03 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Zheng",
"Weijie",
""
],
[
"Doerr",
"Benjamin",
""
]
] |
2312.11419 | Orlando Luongo | Alessio Belfiglio, Orlando Luongo, Stefano Mancini | Superhorizon entanglement from inflationary particle production | 11 pages, 2 figures | Phys.Rev.D 109 (2024) 12, 123520 | 10.1103/PhysRevD.109.123520 | null | gr-qc astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate entanglement generation between the sub- and super-Hubble modes of inflaton fluctuations, in the context of particle production from perturbations during inflation. We consider a large-field inflationary scenario where inflation is driven by a vacuum energy symmetry breaking potential and the scalar inflaton field is nonminimally coupled to spacetime curvature. In particular, we focus on the slow-roll phase, adopting a quasi-de Sitter scale factor to properly account for the presence of perturbations and computing the pair production probability associated to the coupling between the inflaton and spacetime inhomogeneities. The interaction Lagrangian at first order is constructed from inhomogeneities induced by the inflaton dynamics, and the initial Bunch-Davies vacuum state of the field evolves under the action of such Lagrangian. In this framework, we quantify the total amount of entanglement via the von Neumann entropy of the reduced density operator for superhorizon modes, tracing out sub-Hubble degrees of freedom. We then compare these outcomes with entanglement production for quadratic chaotic inflation and for a small-field quadratic hilltop scenario, preserving field-curvature coupling in both cases and pointing out the main differences between large and small-field approaches. We show that the amount of entanglement entropy arising from such geometric production grows rapidly in slow-roll regime and that it is typically higher in large-field scenarios. We also discuss our outcomes in light of recent findings for the squeezing entropy of cosmological perturbations and cubic nonlinearities in de Sitter space. | [
{
"version": "v1",
"created": "Mon, 18 Dec 2023 18:21:06 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Belfiglio",
"Alessio",
""
],
[
"Luongo",
"Orlando",
""
],
[
"Mancini",
"Stefano",
""
]
] |
2312.11651 | Fadi Al Machot | Fadi Al Machot | Bridging Logic and Learning: A Neural-Symbolic Approach for Enhanced Reasoning in Neural Models (ASPER) | null | null | 10.1007/978-3-031-89274-5_5 | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Neural-symbolic learning, an intersection of neural networks and symbolic reasoning, aims to blend neural networks' learning capabilities with symbolic AI's interpretability and reasoning. This paper introduces an approach designed to improve the performance of neural models in learning reasoning tasks. It achieves this by integrating Answer Set Programming (ASP) solvers and domain-specific expertise, which is an approach that diverges from traditional complex neural-symbolic models. In this paper, a shallow artificial neural network (ANN) is specifically trained to solve Sudoku puzzles with minimal training data. The model has a unique loss function that integrates losses calculated using the ASP solver outputs, effectively enhancing its training efficiency. Most notably, the model shows a significant improvement in solving Sudoku puzzles using only 12 puzzles for training and testing without hyperparameter tuning. This advancement indicates that the model's enhanced reasoning capabilities have practical applications, extending well beyond Sudoku puzzles to potentially include a variety of other domains. The code can be found on GitHub: https://github.com/Fadi2200/ASPEN. | [
{
"version": "v1",
"created": "Mon, 18 Dec 2023 19:06:00 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Machot",
"Fadi Al",
""
]
] |
2312.13100 | Fadi Al Machot | William Heyden, Habib Ullah, M. Salman Siddiqui, Fadi Al Machot | SEER-ZSL: Semantic Encoder-Enhanced Representations for Generalized Zero-Shot Learning | null | null | 10.1109/WACVW65960.2025.00009 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Zero-Shot Learning (ZSL) presents the challenge of identifying categories not seen during training. This task is crucial in domains where it is costly, prohibited, or simply not feasible to collect training data. ZSL depends on a mapping between the visual space and available semantic information. Prior works learn a mapping between spaces that can be exploited during inference. We contend, however, that the disparity between meticulously curated semantic spaces and the inherently noisy nature of real-world data remains a substantial and unresolved challenge. In this paper, we address this by introducing a Semantic Encoder-Enhanced Representations for Zero-Shot Learning (SEER-ZSL). We propose a hybrid strategy to address the generalization gap. First, we aim to distill meaningful semantic information using a probabilistic encoder, enhancing the semantic consistency and robustness. Second, we distill the visual space by exploiting the learned data distribution through an adversarially trained generator. Finally, we align the distilled information, enabling a mapping of unseen categories onto the true data manifold. We demonstrate empirically that this approach yields a model that outperforms the state-of-the-art benchmarks in terms of both generalization and benchmarks across diverse settings with small, medium, and large datasets. The complete code is available on GitHub. | [
{
"version": "v1",
"created": "Wed, 20 Dec 2023 15:18:51 GMT"
},
{
"version": "v2",
"created": "Mon, 06 Jan 2025 11:15:11 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Heyden",
"William",
""
],
[
"Ullah",
"Habib",
""
],
[
"Siddiqui",
"M. Salman",
""
],
[
"Machot",
"Fadi Al",
""
]
] |
2312.13316 | Rongsheng Wang | Rongsheng Wang, Qingsong Yao, Zihang Jiang, Haoran Lai, Zhiyang He, Xiaodong Tao, S.Kevin Zhou | ECAMP: Entity-centered Context-aware Medical Vision Language Pre-training | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Despite significant advancements in medical vision-language pre-training, existing methods have largely overlooked the inherent linguistic complexity and imbalanced isssue within medical reports, as well as the complex cross-modality contextual relationships between texts and images. To close this gap, we propose a novel Entity-centered Context-aware Medical Vision-language Pre-training (ECAMP) framework, which establishes a more entity-centered, context-sensitive, and balanced understanding of medical reports to effectively pre-train the vision encoder. We first distill entity-centered context from medical reports utilizing large language models, enabling ECAMP to draw more precise supervision from the text modality. By further incorporating entity-aware re-balanced factor and descriptor masking strategies into masked languange modeling, ECAMP significantly enhances the knowledge of entities within the reports. A context-guided super-resolution task is proposed alongside a multi-scale context fusion design to improve the semantic integration of both coarse and fine-level image representations, which prompts better performance for multi-scale downstream applications. ECAMP integrates these innovations together, leading to significant performance leaps over current state-of-the-art methods and establish a new standard for cross-modality pre-training in medical imaging. The effectiveness of ECAMP is demonstrated by extensive experiments on various domains and organs, which achieves cutting-edge results on multiple tasks including classification, segmentation, and detection across 5 public chest X-ray and 4 fundoscopy datasets respectively. | [
{
"version": "v1",
"created": "Wed, 20 Dec 2023 11:00:54 GMT"
},
{
"version": "v2",
"created": "Mon, 18 Mar 2024 10:54:03 GMT"
},
{
"version": "v3",
"created": "Tue, 19 Mar 2024 11:01:35 GMT"
},
{
"version": "v4",
"created": "Sat, 07 Jun 2025 09:04:59 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Wang",
"Rongsheng",
""
],
[
"Yao",
"Qingsong",
""
],
[
"Jiang",
"Zihang",
""
],
[
"Lai",
"Haoran",
""
],
[
"He",
"Zhiyang",
""
],
[
"Tao",
"Xiaodong",
""
],
[
"Zhou",
"S. Kevin",
""
]
] |
2312.14768 | Guido Giachetti | Guido Giachetti and Nicol\`o Defenu | On the Conditions for a Quantum Violent Relaxation | 6 + 6 pages, 2+2 figures | Phys. Rev. B 111, 214301, 2025 | 10.1103/PhysRevB.111.214301 | null | quant-ph cond-mat.stat-mech | http://creativecommons.org/licenses/by/4.0/ | In general, classical fully-connected systems are known to undergo violent relaxation. This phenomenon refers to the relaxation of observables to stationary, non-thermal, values on a finite timescale, despite their long-time dynamics being dominated by mean-field effects in the thermodynamic limit. Here, we analyze the ``quantum" violent relaxation by studying the dynamics of generic many-body systems with two-body, all-to-all, interactions in the thermodynamic limit. We show that, in order for violent relaxation to occur very specific conditions on the spectrum of the mean-field effective Hamiltonian have to be met. These conditions are hardly met and ``quantum" violent relaxation is observed rarely with respect to its classical counterpart. Our predictions are validated by the study of a spin model which, depending on the value of the coupling, shows a transition between violent-relaxation and a generic prethermal phase. We also analyze a spin version of the quantum Hamiltonian-Mean-Field model, which is shown not to exhibit violent-relaxation. Finally, we discuss how the violent-relaxation picture emerges back in the classical limit. Our results demonstrate how, even in the mean-field regime, quantum effects have a rather dramatic impact on the dynamics, paving the way to a better understanding of light-matter coupled systems. | [
{
"version": "v1",
"created": "Fri, 22 Dec 2023 15:33:26 GMT"
},
{
"version": "v2",
"created": "Sun, 14 Jan 2024 14:06:08 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Giachetti",
"Guido",
""
],
[
"Defenu",
"Nicolò",
""
]
] |
2401.00793 | Jinglong Luo | Jinglong Luo, Yehong Zhang, Zhuo Zhang, Jiaqi Zhang, Xin Mu, Hui Wang, Yue Yu, Zenglin Xu | SecFormer: Fast and Accurate Privacy-Preserving Inference for Transformer Models via SMPC | ACL 2024 | null | null | null | cs.LG cs.CL cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the growing use of Transformer models hosted on cloud platforms to offer inference services, privacy concerns are escalating, especially concerning sensitive data like investment plans and bank account details. Secure Multi-Party Computing (SMPC) emerges as a promising solution to protect the privacy of inference data and model parameters. However, the application of SMPC in Privacy-Preserving Inference (PPI) for Transformer models often leads to considerable slowdowns or declines in performance. This is largely due to the multitude of nonlinear operations in the Transformer architecture, which are not well-suited to SMPC and difficult to circumvent or optimize effectively. To address this concern, we introduce a comprehensive PPI framework called SecFormer to achieve fast and accurate PPI for Transformer models. We successfully eliminate the high-cost exponential and maximum operations in PPI without sacrificing model performance and develop a suite of efficient SMPC protocols by employing suitable numerical computation methods to boost other complex nonlinear functions in PPI, including GeLU, LayerNorm, and a redesigned Softmax. Our extensive experiments reveal that SecFormer outperforms MPCFormer in performance, showing improvements of $3.4\%$ and $24.7\%$ for BERT$_{\text{BASE}}$ and BERT$_{\text{LARGE}}$, respectively. In terms of efficiency, SecFormer is 3.57 and 3.58 times faster than PUMA for BERT$_{\text{BASE}}$ and BERT$_{\text{LARGE}}$, demonstrating its effectiveness and speed. | [
{
"version": "v1",
"created": "Mon, 01 Jan 2024 15:40:35 GMT"
},
{
"version": "v2",
"created": "Sat, 06 Jan 2024 10:05:23 GMT"
},
{
"version": "v3",
"created": "Thu, 06 Jun 2024 05:22:44 GMT"
},
{
"version": "v4",
"created": "Sat, 14 Dec 2024 02:42:10 GMT"
},
{
"version": "v5",
"created": "Mon, 09 Jun 2025 07:49:24 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Luo",
"Jinglong",
""
],
[
"Zhang",
"Yehong",
""
],
[
"Zhang",
"Zhuo",
""
],
[
"Zhang",
"Jiaqi",
""
],
[
"Mu",
"Xin",
""
],
[
"Wang",
"Hui",
""
],
[
"Yu",
"Yue",
""
],
[
"Xu",
"Zenglin",
""
]
] |
2401.05524 | Scott Melville | Scott Melville | Causality and quasi-normal modes in the GREFT | 44 pages, 5 figures, 1 table | Eur.Phys.J.Plus 139 (2024) 8, 725 | 10.1140/epjp/s13360-024-05520-5 | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The General Relativity Effective Field Theory (GREFT) introduces higher-derivative interactions to parameterise the gravitational effects of massive degrees of freedom which are too heavy to be probed directly. The coefficients of these interactions have recently been constrained using causality: both from the analytic structure of 4-point graviton scattering and the time delay of gravitational waves on a black hole background. In this work, causality is used to constrain the quasi-normal mode spectrum of GREFT black holes. Demanding that quasi-normal mode perturbations decay faster in the GREFT than in General Relativity -- a new kind of causality condition which stems from the analytic structure of 2-point functions on a black hole background -- leads to further constraints on the GREFT coefficients. The causality constraints and compact expressions for the GREFT quasi-normal mode frequencies presented here will inform future parameterised gravitational waveforms, and the observational prospects for gravitational wave observatories are briefly discussed. | [
{
"version": "v1",
"created": "Wed, 10 Jan 2024 19:42:46 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Melville",
"Scott",
""
]
] |
2401.10969 | Gianluca Aguzzi | Gianluca Aguzzi, Roberto Casadei, Mirko Viroli | MacroSwarm: A Field-based Compositional Framework for Swarm Programming | null | null | null | null | cs.AI cs.LO cs.SE | http://creativecommons.org/licenses/by/4.0/ | Swarm behaviour engineering is an area of research that seeks to investigate methods and techniques for coordinating computation and action within groups of simple agents to achieve complex global goals like pattern formation, collective movement, clustering, and distributed sensing. Despite recent progress in the analysis and engineering of swarms (of drones, robots, vehicles), there is still a need for general design and implementation methods and tools that can be used to define complex swarm behaviour in a principled way. To contribute to this quest, this article proposes a new field-based coordination approach, called MacroSwarm, to design and program swarm behaviour in terms of reusable and fully composable functional blocks embedding collective computation and coordination. Based on the macroprogramming paradigm of aggregate computing, MacroSwarm builds on the idea of expressing each swarm behaviour block as a pure function, mapping sensing fields into actuation goal fields, e.g., including movement vectors. In order to demonstrate the expressiveness, compositionality, and practicality of MacroSwarm as a framework for swarm programming, we perform a variety of simulations covering common patterns of flocking, pattern formation, and collective decision-making. The implications of the inherent self-stabilisation properties of field-based computations in MacroSwarm are discussed, which formally guarantee some resilience properties and guided the design of the library. | [
{
"version": "v1",
"created": "Fri, 19 Jan 2024 16:32:02 GMT"
},
{
"version": "v2",
"created": "Mon, 11 Nov 2024 09:51:28 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 14:39:12 GMT"
},
{
"version": "v4",
"created": "Mon, 09 Jun 2025 10:14:13 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Aguzzi",
"Gianluca",
""
],
[
"Casadei",
"Roberto",
""
],
[
"Viroli",
"Mirko",
""
]
] |
2401.11792 | Zj Tang | Zuojin Tang, Xiaoyu Chen, Yongqiang Li, Jianyu Chen | Efficient and Generalized end-to-end Autonomous Driving System with Latent Deep Reinforcement Learning and Demonstrations | Accepted by ECML PKDD 2025 (Research Track) | null | null | null | cs.RO cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An intelligent driving system should dynamically formulate appropriate driving strategies based on the current environment and vehicle status while ensuring system security and reliability. However, methods based on reinforcement learning and imitation learning often suffer from high sample complexity, poor generalization, and low safety. To address these challenges, this paper introduces an efficient and generalized end-to-end autonomous driving system (EGADS) for complex and varied scenarios. The RL agent in our EGADS combines variational inference with normalizing flows, which are independent of distribution assumptions. This combination allows the agent to capture historical information relevant to driving in latent space effectively, thereby significantly reducing sample complexity. Additionally, we enhance safety by formulating robust safety constraints and improve generalization and performance by integrating RL with expert demonstrations. Experimental results demonstrate that, compared to existing methods, EGADS significantly reduces sample complexity, greatly improves safety performance, and exhibits strong generalization capabilities in complex urban scenarios. Particularly, we contributed an expert dataset collected through human expert steering wheel control, specifically using the G29 steering wheel. | [
{
"version": "v1",
"created": "Mon, 22 Jan 2024 09:44:16 GMT"
},
{
"version": "v2",
"created": "Tue, 23 Jan 2024 02:53:58 GMT"
},
{
"version": "v3",
"created": "Wed, 24 Jan 2024 09:04:03 GMT"
},
{
"version": "v4",
"created": "Mon, 29 Jan 2024 01:50:12 GMT"
},
{
"version": "v5",
"created": "Thu, 14 Mar 2024 07:47:03 GMT"
},
{
"version": "v6",
"created": "Sun, 16 Jun 2024 12:48:53 GMT"
},
{
"version": "v7",
"created": "Mon, 09 Jun 2025 15:02:33 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Tang",
"Zuojin",
""
],
[
"Chen",
"Xiaoyu",
""
],
[
"Li",
"Yongqiang",
""
],
[
"Chen",
"Jianyu",
""
]
] |
2401.12018 | Aaron Hurst | Aaron Hurst and Daniel E. Lucani and Qi Zhang | PairwiseHist: Fast, Accurate and Space-Efficient Approximate Query Processing with Data Compression | null | null | 10.14778/3648160.3648181 | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Exponential growth in data collection is creating significant challenges for data storage and analytics latency.Approximate Query Processing (AQP) has long been touted as a solution for accelerating analytics on large datasets, however, there is still room for improvement across all key performance criteria. In this paper, we propose a novel histogram-based data synopsis called PairwiseHist that uses recursive hypothesis testing to ensure accurate histograms and can be built on top of data compressed using Generalized Deduplication (GD). We thus show that GD data compression can contribute to AQP. Compared to state-of-the-art AQP approaches, PairwiseHist achieves better performance across all key metrics, including 2.6$ \times $ higher accuracy, 3.5$ \times $ lower latency, 24$ \times $ smaller synopses and 1.5--4$ \times $ faster construction time. | [
{
"version": "v1",
"created": "Mon, 22 Jan 2024 15:04:46 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Hurst",
"Aaron",
""
],
[
"Lucani",
"Daniel E.",
""
],
[
"Zhang",
"Qi",
""
]
] |
2401.13577 | Elvira Lupoian | Elvira Lupoian and James Rawson | Three-Torsion Subgroups and Wild Conductor Exponents of Plane Quartics | Major changes: change of title, re-organised structure and added implementation. Comments are welcome! | null | null | null | math.NT | http://creativecommons.org/licenses/by-sa/4.0/ | In this paper we give an algorithm to find the 3-torsion subgroup of the Jacobian of a smooth plane quartic curve with a marked rational point. We describe $3-$torsion points in terms of cubics which triply intersect the curve, and use this to define a system of equations whose solution set corresponds to the coefficients of these cubics. We compute the points of this zero-dimensional, degree $728$ scheme first by approximation, using homotopy continuation and Newton-Raphson, and then using continued fractions to obtain accurate expressions for these points. We describe how the Galois structure of the field of definition of the $3$-torsion subgroup can be used to compute local wild conductor exponents, including at $p=2$. | [
{
"version": "v1",
"created": "Wed, 24 Jan 2024 16:41:19 GMT"
},
{
"version": "v2",
"created": "Sat, 03 May 2025 19:52:23 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 16:01:28 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Lupoian",
"Elvira",
""
],
[
"Rawson",
"James",
""
]
] |
2401.16185 | Yuqiang Sun | Yuqiang Sun and Daoyuan Wu and Yue Xue and Han Liu and Wei Ma and Lyuye Zhang and Yang Liu and Yingjiu Li | LLM4Vuln: A Unified Evaluation Framework for Decoupling and Enhancing LLMs' Vulnerability Reasoning | This is a technical report by Nanyang Technological University. Updated to support Solidity, Java and C/C++ | null | null | null | cs.CR cs.AI cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) have demonstrated significant potential in various tasks, including those requiring human-level intelligence, such as vulnerability detection. However, recent efforts to use LLMs for vulnerability detection remain preliminary, as they lack a deep understanding of whether a subject LLM's vulnerability reasoning capability stems from the model itself or from external aids such as knowledge retrieval and tooling support.
In this paper, we aim to decouple LLMs' vulnerability reasoning from other capabilities, such as vulnerability knowledge adoption, context information retrieval, and advanced prompt schemes. We introduce LLM4Vuln, a unified evaluation framework that separates and assesses LLMs' vulnerability reasoning capabilities and examines improvements when combined with other enhancements.
To support this evaluation, we construct UniVul, the first benchmark that provides retrievable knowledge and context-supplementable code across three representative programming languages: Solidity, Java, and C/C++. Using LLM4Vuln and UniVul, we test six representative LLMs (GPT-4.1, Phi-3, Llama-3, o4-mini, DeepSeek-R1, and QwQ-32B) for 147 ground-truth vulnerabilities and 147 non-vulnerable cases in 3,528 controlled scenarios. Our findings reveal the varying impacts of knowledge enhancement, context supplementation, and prompt schemes. We also identify 14 zero-day vulnerabilities in four pilot bug bounty programs, resulting in $3,576 in bounties. | [
{
"version": "v1",
"created": "Mon, 29 Jan 2024 14:32:27 GMT"
},
{
"version": "v2",
"created": "Thu, 05 Sep 2024 15:03:51 GMT"
},
{
"version": "v3",
"created": "Mon, 13 Jan 2025 06:10:24 GMT"
},
{
"version": "v4",
"created": "Sat, 07 Jun 2025 17:03:11 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Sun",
"Yuqiang",
""
],
[
"Wu",
"Daoyuan",
""
],
[
"Xue",
"Yue",
""
],
[
"Liu",
"Han",
""
],
[
"Ma",
"Wei",
""
],
[
"Zhang",
"Lyuye",
""
],
[
"Liu",
"Yang",
""
],
[
"Li",
"Yingjiu",
""
]
] |
2401.16508 | Jack Davies | Christian Carrick and Jack Morgan Davies | A synthetic approach to detecting $v_1$-periodic families | 53 pages and 16 spectral sequence diagrams. Comments are always welcome. v2 adds a proof of Wood's theorem and refines assumptions. Accepted version | null | null | null | math.AT math.KT | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We provide a simple proof that the unit map from the sphere spectrum to the connective image-of-$J$ spectrum $\mathrm{j}$ is surjective on homotopy groups. This is achieved using a novel $t$-structure on the category of $E$-synthetic spectra and a specific construction of $\mathbf{F}_p$- and BP-synthetic lifts of $\mathrm{j}$. These synthetic lifts then easily produce modified Adams and Adams--Novikov spectral sequences for $\mathrm{j}$ which we use the prove the above detection statement, all without ever calculating $\mathbf{F}_p$- or BP-homology nor the associated Ext groups. | [
{
"version": "v1",
"created": "Mon, 29 Jan 2024 19:25:56 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 10:31:10 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Carrick",
"Christian",
""
],
[
"Davies",
"Jack Morgan",
""
]
] |
2401.17119 | Silv\`ere Gangloff | Silv\`ere Gangloff and Alonso N\'u\~nez | The topological structure of isolated points in the space of $\mathbb{Z}^d$-shifts | null | null | null | null | math.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | R. Pavlov and S. Schmieding provided recently some results about generic $\mathbb{Z}$-shifts, which rely mainly on an original theorem stating that isolated points form a residual set in the space of $\mathbb{Z}$-shifts such that all other residual sets must contain it. As a direction for further research, they pointed towards genericity in the space of G-shifts, where G is a finitely generated group. In the present text, we approach this for the case of $\mathbb{Z}^d$-shifts, where $d \ge 2$. As it is usual, multidimensional dynamical systems are much more difficult to understand. In light of the result of R. Pavlov and S. Schmieding, it is natural to begin with a better understanding of isolated points. We prove here a characterization of such points in the space of $\mathbb{Z}^d$-shifts, in terms of the natural notion of maximal subsystems which we also introduce in this article. From this characterization we recover the result of R. Pavlov and S. Schmieding's for $\mathbb{Z}$-shifts. We also prove a series of results which exploit this notion. In particular some transitivity-like properties can be related to the number of maximal subsystems. Furthermore, we show that the Cantor-Bendixon rank of the space of $\mathbb{Z}^d$-shifts is infinite for $d > 1$, while for $d = 1$ is known to be equal to one. | [
{
"version": "v1",
"created": "Tue, 30 Jan 2024 15:53:13 GMT"
},
{
"version": "v2",
"created": "Thu, 01 Feb 2024 11:50:28 GMT"
},
{
"version": "v3",
"created": "Mon, 22 Apr 2024 20:04:08 GMT"
},
{
"version": "v4",
"created": "Sat, 11 May 2024 12:21:47 GMT"
},
{
"version": "v5",
"created": "Fri, 22 Nov 2024 21:44:40 GMT"
},
{
"version": "v6",
"created": "Sat, 22 Feb 2025 12:54:40 GMT"
},
{
"version": "v7",
"created": "Mon, 09 Jun 2025 12:31:23 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Gangloff",
"Silvère",
""
],
[
"Núñez",
"Alonso",
""
]
] |
2401.17852 | Jan Harold Alcantara | Jan Harold Alcantara, Akiko Takeda | Theoretical smoothing frameworks for nonsmooth simple bilevel problems | 32 pages | null | null | null | math.OC | http://creativecommons.org/licenses/by/4.0/ | Bilevel programming has recently received a great deal of attention due to its abundant applications in many areas. The optimal value function approach provides a useful reformulation of the bilevel problem, but its utility is often limited due to the nonsmoothness of the value function even in cases when the associated lower-level function is smooth. In this paper, we present two smoothing strategies for the value function associated with lower-level functions that are not necessarily smooth but are Lipschitz continuous. The first method employs quadratic regularization for partially convex lower-level functions, while the second utilizes entropic regularization for general lower-level objective functions. Meanwhile, the property known as gradient consistency is crucial in ensuring that a designed smoothing algorithm is globally subsequentially convergent to stationary points of the value function reformulation. With this motivation, we prove that the proposed smooth approximations satisfy the gradient consistent property under certain conditions on the lower-level function. | [
{
"version": "v1",
"created": "Wed, 31 Jan 2024 14:14:11 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 05:23:36 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Alcantara",
"Jan Harold",
""
],
[
"Takeda",
"Akiko",
""
]
] |
2402.02793 | Martin Hanke | Martin Hanke | On the shape derivative of polygonal inclusions in the conductivity problem | null | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | We consider the conductivity problem for a homogeneous body with an inclusion of a different, but known, conductivity. Our interest concerns the associated shape derivative, i.e., the derivative of the corresponding electrostatic potential with respect to the shape of the inclusion. For a smooth inclusion it is known that the shape derivative is the solution of a specific inhomogeneous transmission problem. We show that this characterization of the shape derivative is also valid when the inclusion is a polygonal domain, but due to singularities at the vertices of the polygon, the shape derivative fails to belong to $H^1$ in this case. | [
{
"version": "v1",
"created": "Mon, 05 Feb 2024 08:01:23 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 09:35:39 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Hanke",
"Martin",
""
]
] |
2402.03275 | Lorenzo Mercuri | Lorenzo Mercuri, Andrea Perchiazzo and Edit Rroji | Simulation thinning algorithm for a CARMA(p,q)-Hawkes model | The working paper is now part of an upcoming article in Quantitative Finance titled "Option Pricing with a Compound CARMA(p,q)-Hawkes Process" (arXiv:2412.15172; see Section 4: Simulation Algorithm). To avoid any potential confusion for readers, we kindly request the withdrawal of the working paper | null | null | null | stat.CO | http://creativecommons.org/licenses/by/4.0/ | This paper presents an algorithm for the simulation of Hawkes-type processes where the intensity is expressed in terms of a continuous-time autoregressive moving average model. We identify upper bounds for both the univariate and the multivariate intensity functions that are used to develop simulation algorithms based on the thinning technique. | [
{
"version": "v1",
"created": "Mon, 05 Feb 2024 18:33:15 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 10:16:27 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Mercuri",
"Lorenzo",
""
],
[
"Perchiazzo",
"Andrea",
""
],
[
"Rroji",
"Edit",
""
]
] |
2402.03292 | Huy Nguyen Quang | Quang-Huy Nguyen, Jin Peng Zhou, Zhenzhen Liu, Khanh-Huyen Bui, Kilian Q. Weinberger, Wei-Lun Chao, Dung D. Le | Detecting Out-of-Distribution Objects through Class-Conditioned Inpainting | null | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recent object detectors have achieved impressive accuracy in identifying objects seen during training. However, real-world deployment often introduces novel and unexpected objects, referred to as out-of-distribution (OOD) objects, posing significant challenges to model trustworthiness. Modern object detectors are typically overconfident, making it unreliable to use their predictions alone for OOD detection. To address this, we propose leveraging an auxiliary model as a complementary solution. Specifically, we utilize an off-the-shelf text-to-image generative model, such as Stable Diffusion, which is trained with objective functions distinct from those of discriminative object detectors. We hypothesize that this fundamental difference enables the detection of OOD objects by measuring inconsistencies between the models. Concretely, for a given detected object bounding box and its predicted in-distribution class label, we perform class-conditioned inpainting on the image with the object removed. If the object is OOD, the inpainted image is likely to deviate significantly from the original, making the reconstruction error a robust indicator of OOD status. Extensive experiments demonstrate that our approach consistently surpasses existing zero-shot and non-zero-shot OOD detection methods, establishing a robust framework for enhancing object detection systems in dynamic environments. | [
{
"version": "v1",
"created": "Mon, 05 Feb 2024 18:50:27 GMT"
},
{
"version": "v2",
"created": "Wed, 07 Feb 2024 03:21:57 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 14:06:04 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Nguyen",
"Quang-Huy",
""
],
[
"Zhou",
"Jin Peng",
""
],
[
"Liu",
"Zhenzhen",
""
],
[
"Bui",
"Khanh-Huyen",
""
],
[
"Weinberger",
"Kilian Q.",
""
],
[
"Chao",
"Wei-Lun",
""
],
[
"Le",
"Dung D.",
""
]
] |
2402.04062 | Xingyue Huang | Xingyue Huang, Miguel Romero Orth, Pablo Barcel\'o, Michael M. Bronstein, \.Ismail \.Ilkan Ceylan | Link Prediction with Relational Hypergraphs | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Link prediction with knowledge graphs has been thoroughly studied in graph machine learning, leading to a rich landscape of graph neural network architectures with successful applications. Nonetheless, it remains challenging to transfer the success of these architectures to relational hypergraphs, where the task of link prediction is over $k$-ary relations, which is substantially harder than link prediction with knowledge graphs. In this paper, we propose a framework for link prediction with relational hypergraphs, unlocking applications of graph neural networks to fully relational structures. Theoretically, we conduct a thorough analysis of the expressive power of the resulting model architectures via corresponding relational Weisfeiler-Leman algorithms and also via logical expressiveness. Empirically, we validate the power of the proposed model architectures on various relational hypergraph benchmarks. The resulting model architectures substantially outperform every baseline for inductive link prediction, and lead to state-of-the-art results for transductive link prediction. | [
{
"version": "v1",
"created": "Tue, 06 Feb 2024 15:05:40 GMT"
},
{
"version": "v2",
"created": "Thu, 23 May 2024 15:37:16 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 10:08:14 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Huang",
"Xingyue",
""
],
[
"Orth",
"Miguel Romero",
""
],
[
"Barceló",
"Pablo",
""
],
[
"Bronstein",
"Michael M.",
""
],
[
"Ceylan",
"İsmail İlkan",
""
]
] |
2402.05100 | Kengo Kato | Kengo Kato | Large deviations for dynamical Schr\"{o}dinger problems | 23 pages | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | We establish large deviations for dynamical Schr\"{o}dinger problems driven by perturbed Brownian motions when the noise parameter tends to zero. Our results show that Schr\"{o}dinger bridges charge exponentially small masses outside the support of the limiting law that agrees with the optimal solution to the dynamical Monge-Kantorovich optimal transport problem. Our proofs build on mixture representations of Schr\"{o}dinger bridges and establishing exponential continuity of Brownian bridges with respect to the initial and terminal points. | [
{
"version": "v1",
"created": "Wed, 07 Feb 2024 18:55:15 GMT"
},
{
"version": "v2",
"created": "Tue, 13 Feb 2024 04:31:08 GMT"
},
{
"version": "v3",
"created": "Sat, 07 Jun 2025 16:41:04 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Kato",
"Kengo",
""
]
] |
2402.05430 | Aseem Rajan Kshirsagar | Aseem Rajan Kshirsagar, Sven Reichardt | Flipping of electronic spins in BiFeO$_3$ via chiral $d-d$ excitations | null | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | BiFeO$_3$ is a multiferroic material featuring ferroelectricity and noncollinear antiferromagnetism. Definitive and efficient control of the characteristic spin texture of BiFeO$_3$ is attractive for emerging quantum devices. In this regard, crystal-field $d\rightarrow d$ excitations localized on Fe atomic sites in BiFeO$_3$ provide an avenue for manipulation of the spin texture as they induce a complex interplay among the spin, charge, and lattice degrees of freedom. In this work, the \textit{ab initio} \textit{GW}-BSE method is used to characterize these excitations within an excitonic picture. We find that the $d-d$ transitions appear as strongly bound, chiral, spin-flip excitons deep within the electronic band gap as a result of the intricate competition between the lattice potential, the antiferromagnetic ordering, the spin-orbit coupling, and the electron-hole interaction. Most crucially, these excitons are composed of electron-hole pairs with opposite spins that constitute almost all of their $\pm \hbar$ total angular momentum. These excitons of specific angular momentum can be selectively excited using circularly polarized light, consequently modulating the local magnetic moment. | [
{
"version": "v1",
"created": "Thu, 08 Feb 2024 05:56:23 GMT"
},
{
"version": "v2",
"created": "Wed, 21 Feb 2024 16:41:36 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 17:04:38 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Kshirsagar",
"Aseem Rajan",
""
],
[
"Reichardt",
"Sven",
""
]
] |
2402.05954 | Sijun Xia | Jianming Lv, Sijun Xia, Depin Liang, Wei Chen | EasyFS: an Efficient Model-free Feature Selection Framework via Elastic Transformation of Features | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional model-free feature selection methods treat each feature independently while disregarding the interrelationships among features, which leads to relatively poor performance compared with the model-aware methods. To address this challenge, we propose an efficient model-free feature selection framework via elastic expansion and compression of the features, namely EasyFS, to achieve better performance than state-of-the-art model-aware methods while sharing the characters of efficiency and flexibility with the existing model-free methods. In particular, EasyFS expands the feature space by using the random non-linear projection network to achieve the non-linear combinations of the original features, so as to model the interrelationships among the features and discover most correlated features. Meanwhile, a novel redundancy measurement based on the change of coding rate is proposed for efficient filtering of redundant features. Comprehensive experiments on 21 different datasets show that EasyFS outperforms state-of-the art methods up to 10.9\% in the regression tasks and 5.7\% in the classification tasks while saving more than 94\% of the time. | [
{
"version": "v1",
"created": "Sun, 04 Feb 2024 09:25:07 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 13:43:16 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Lv",
"Jianming",
""
],
[
"Xia",
"Sijun",
""
],
[
"Liang",
"Depin",
""
],
[
"Chen",
"Wei",
""
]
] |
2402.07660 | Lei Yu | Lei Yu | R\'enyi Resolvability, Noise Stability, and Anti-contractivity | To be published in IEEE Transactions on Information Theory. For saving space, the proofs of Lemma 11 (One-Shot Bounds) and Theorem 5 (Exponential Behavior of Typical Codes) are omitted in the published version | null | null | null | cs.IT math.FA math.IT math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper investigates three closely related topics -- R\'enyi resolvability, noise stability, and anti-contractivity. The R\'enyi resolvability problem refers to approximating a target output distribution of a given channel in the R\'enyi divergence when the input is set to a function of a given uniform random variable. This problem for the R\'enyi parameter in $(0,2]\cup\{\infty\}$ was first studied by the present author and Tan in 2019. In the present paper, we provide a complete solution to this problem for the R\'enyi parameter in the entire range $\mathbb{R}\cup\{\pm\infty\}$. We then connect the R\'enyi resolvability problem to the noise stability problem, by observing that maximizing or minimizing the $q$-stability of a set is equivalent to a variant of the R\'enyi resolvability problem. By such a connection, we provide sharp dimension-free bounds on the $q$-stability. We lastly relate the noise stability problem to the anti-contractivity of a Markov operator (i.e., conditional expectation operator), where the terminology ``anti-contractivity'' introduced by us refers to as the opposite property of the well-known contractivity/hyercontractivity. We derive sharp dimension-free anti-contractivity inequalities. All of the results in this paper are evaluated for binary distributions. Our proofs in this paper are mainly based on the method of types, especially strengthened versions of packing-covering lemmas. | [
{
"version": "v1",
"created": "Mon, 12 Feb 2024 14:06:24 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Feb 2024 14:07:11 GMT"
},
{
"version": "v3",
"created": "Fri, 21 Jun 2024 15:21:54 GMT"
},
{
"version": "v4",
"created": "Mon, 03 Feb 2025 13:36:35 GMT"
},
{
"version": "v5",
"created": "Sat, 07 Jun 2025 13:25:17 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Yu",
"Lei",
""
]
] |
2402.10500 | Nirjhar Das | Nirjhar Das, Souradip Chakraborty, Aldo Pacchiano, Sayak Ray Chowdhury | Active Preference Optimization for Sample Efficient RLHF | Accepted at ECML-PKDD 2025. Camera ready version | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Large Language Models (LLMs) aligned using Reinforcement Learning from Human Feedback (RLHF) have shown remarkable generation abilities in numerous tasks. However, collecting high-quality human preferences creates costly bottlenecks in practical deployments, and hence, training data are often budgeted. In these scenarios, it is crucial to collect training data (e.g., contexts, a pair of generations for each context, and a preference indicating which generation is better) carefully, yet most of the existing methods sample contexts uniformly at random from a given collection. Given this, under the Bradley-Terry-Luce preference model and with a small budget of training data, we show that uniform sampling of contexts could lead to a policy (i.e., an aligned model) that suffers a constant sub-optimality gap from the optimal policy. This highlights the need for an adaptive context sampling strategy for effective alignment under a small sample budget. To address this, we reformulate RLHF within the contextual preference bandit framework, treating generations as actions, and give a nearly complete characterization of the sub-optimality gap in terms of both lower and upper bounds. First, when the action set is a $d$-dimensional hypercube and the number of samples is $T$, we show an $\Omega(d/\sqrt{T})$ lower bound. Next, we propose an algorithm, $\textit{Active Preference Optimization}$ ($\texttt{APO}$), that iteratively collects preferences for the most uncertain contexts. We show that the sub-optimality gap of the policy learned via $\texttt{APO}$ matches the lower bound up to a log factor and a non-linearity constant. Finally, we perform experiments on practical datasets to validate $\texttt{APO}$'s efficacy over existing methods, establishing it as a sample-efficient and cost-effective solution for LLM alignment. | [
{
"version": "v1",
"created": "Fri, 16 Feb 2024 08:19:34 GMT"
},
{
"version": "v2",
"created": "Wed, 05 Jun 2024 15:10:08 GMT"
},
{
"version": "v3",
"created": "Sat, 07 Jun 2025 16:03:51 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Das",
"Nirjhar",
""
],
[
"Chakraborty",
"Souradip",
""
],
[
"Pacchiano",
"Aldo",
""
],
[
"Chowdhury",
"Sayak Ray",
""
]
] |
2402.10504 | Elad Aigner-Horev | Elad Aigner-Horev, and Daniel Rosenberg, and Roi Weiss | Resilience of Rademacher chaos of low degree | Touch-ups in the introduction | null | null | null | math.PR cs.IT cs.LG math.CO math.IT stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The resilience of a Rademacher chaos is the maximum number of adversarial sign-flips that the chaos can sustain without having its largest atom probability significantly altered. Inspired by probabilistic lower-bound guarantees for the resilience of linear Rademacher chaos, obtained by Bandeira, Ferber, and Kwan (Advances in Mathematics, Vol. $319$, $2017$), we provide probabilistic lower-bound guarantees for the resilience of Rademacher chaos of arbitrary yet sufficiently low degree.
Our main results distinguish between Rademacher chaos of order two and those of higher order. In that, our first main result pertains to the resilience of decoupled bilinear Rademacher forms where different asymptotic behaviour is observed for sparse and dense matrices. For our second main result, we bootstrap our first result in order to provide resilience guarantees for quadratic Rademacher chaos. Our third main result, generalises the first and handles the resilience of decoupled Rademacher chaos of arbitrary yet sufficiently low order.
Our results for decoupled Rademacher chaos of order two and that of higher order whilst are established through the same conceptual framework, differ substantially. A difference incurred due to the implementation of the same conceptual argument. The order two result is established using Dudley's maximal inequality for sub-Gaussian processes, the Hanson-Wright inequality, as well as the Kolmogorov-Rogozin inequality. To handle higher order chaos, appeals to Dudley's inequality as well as the Hanson-Wright inequality are replaced with tools suited for random tensors. Appeals to the Hanson-Wright inequality are replaced with appeals to a concentration result for random tensors put forth by Adamczak and Wolff.
Our results are instance-dependent and thus allow for the efficient computation of resilience guarantees provided the order of the chaos is constant. | [
{
"version": "v1",
"created": "Fri, 16 Feb 2024 08:27:55 GMT"
},
{
"version": "v2",
"created": "Mon, 05 Aug 2024 08:22:56 GMT"
},
{
"version": "v3",
"created": "Sat, 15 Mar 2025 18:59:02 GMT"
},
{
"version": "v4",
"created": "Wed, 16 Apr 2025 09:25:06 GMT"
},
{
"version": "v5",
"created": "Mon, 09 Jun 2025 05:45:25 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Aigner-Horev",
"Elad",
""
],
[
"Rosenberg",
"Daniel",
""
],
[
"Weiss",
"Roi",
""
]
] |
2402.10686 | Meiyi Zhu | Meiyi Zhu, Caili Guo, Chunyan Feng, Osvaldo Simeone | On the Impact of Uncertainty and Calibration on Likelihood-Ratio Membership Inference Attacks | 16 pages, 28 figures | null | null | null | cs.IT cs.CR cs.LG eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In a membership inference attack (MIA), an attacker exploits the overconfidence exhibited by typical machine learning models to determine whether a specific data point was used to train a target model. In this paper, we analyze the performance of the likelihood ratio attack (LiRA) within an information-theoretical framework that allows the investigation of the impact of the aleatoric uncertainty in the true data generation process, of the epistemic uncertainty caused by a limited training data set, and of the calibration level of the target model. We compare three different settings, in which the attacker receives decreasingly informative feedback from the target model: confidence vector (CV) disclosure, in which the output probability vector is released; true label confidence (TLC) disclosure, in which only the probability assigned to the true label is made available by the model; and decision set (DS) disclosure, in which an adaptive prediction set is produced as in conformal prediction. We derive bounds on the advantage of an MIA adversary with the aim of offering insights into the impact of uncertainty and calibration on the effectiveness of MIAs. Simulation results demonstrate that the derived analytical bounds predict well the effectiveness of MIAs. | [
{
"version": "v1",
"created": "Fri, 16 Feb 2024 13:41:18 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Aug 2024 16:04:29 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Mar 2025 10:13:46 GMT"
},
{
"version": "v4",
"created": "Tue, 13 May 2025 15:38:09 GMT"
},
{
"version": "v5",
"created": "Mon, 09 Jun 2025 01:29:17 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Zhu",
"Meiyi",
""
],
[
"Guo",
"Caili",
""
],
[
"Feng",
"Chunyan",
""
],
[
"Simeone",
"Osvaldo",
""
]
] |
2402.10793 | David Buterez | David Buterez, Jon Paul Janet, Dino Oglic, Pietro Lio | An end-to-end attention-based approach for learning on graphs | Now published in Nature Communications | Nat Commun 16, 5244 (2025) | 10.1038/s41467-025-60252-z | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | There has been a recent surge in transformer-based architectures for learning on graphs, mainly motivated by attention as an effective learning mechanism and the desire to supersede handcrafted operators characteristic of message passing schemes. However, concerns over their empirical effectiveness, scalability, and complexity of the pre-processing steps have been raised, especially in relation to much simpler graph neural networks that typically perform on par with them across a wide range of benchmarks. To tackle these shortcomings, we consider graphs as sets of edges and propose a purely attention-based approach consisting of an encoder and an attention pooling mechanism. The encoder vertically interleaves masked and vanilla self-attention modules to learn an effective representations of edges, while allowing for tackling possible misspecifications in input graphs. Despite its simplicity, the approach outperforms fine-tuned message passing baselines and recently proposed transformer-based methods on more than 70 node and graph-level tasks, including challenging long-range benchmarks. Moreover, we demonstrate state-of-the-art performance across different tasks, ranging from molecular to vision graphs, and heterophilous node classification. The approach also outperforms graph neural networks and transformers in transfer learning settings, and scales much better than alternatives with a similar performance level or expressive power. | [
{
"version": "v1",
"created": "Fri, 16 Feb 2024 16:20:11 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Dec 2024 15:44:46 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 13:24:10 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Buterez",
"David",
""
],
[
"Janet",
"Jon Paul",
""
],
[
"Oglic",
"Dino",
""
],
[
"Lio",
"Pietro",
""
]
] |
2402.11370 | Sara Fish | Sara Fish, Yannai A. Gonczarowski, Sergiu Hart | Stable Menus of Public Goods: A Matching Problem | Accepted to EC 2025 | null | null | null | cs.GT econ.TH math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study a matching problem between agents and public goods, in settings without monetary transfers. Since goods are public, they have no capacity constraints. There is no exogenously defined budget of goods to be provided. Rather, each provided good must justify its cost by being utilized by sufficiently many agents, leading to strong complementarities in the "preferences" of goods. Furthermore, goods that are in high demand given other already-provided goods must also be provided. The question of the existence of a stable solution (a menu of public goods to be provided) exhibits a rich combinatorial structure. We uncover sufficient conditions and necessary conditions for guaranteeing the existence of a stable solution, and derive both positive and negative results for strategyproof stable matching. | [
{
"version": "v1",
"created": "Sat, 17 Feb 2024 20:13:30 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Jun 2025 20:50:37 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Fish",
"Sara",
""
],
[
"Gonczarowski",
"Yannai A.",
""
],
[
"Hart",
"Sergiu",
""
]
] |
2402.11711 | Yasaman Jafari | Yasaman Jafari, Dheeraj Mekala, Rose Yu, Taylor Berg-Kirkpatrick | MORL-Prompt: An Empirical Analysis of Multi-Objective Reinforcement Learning for Discrete Prompt Optimization | null | null | 10.18653/v1/2024.findings-emnlp.577 | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | RL-based techniques can be employed to search for prompts that, when fed into a target language model, maximize a set of user-specified reward functions. However, in many target applications, the natural reward functions are in tension with one another -- for example, content preservation vs. style matching in style transfer tasks. Current techniques focus on maximizing the average of reward functions, which does not necessarily lead to prompts that achieve balance across rewards -- an issue that has been well-studied in the multi-objective and robust optimization literature. In this paper, we conduct an empirical comparison of several existing multi-objective optimization techniques adapted to this new setting: RL-based discrete prompt optimization. We compare two methods optimizing the volume of the Pareto reward surface and one method that chooses an update direction that benefits all rewards simultaneously. We evaluate performance on two NLP tasks: style transfer and machine translation, each using three competing reward functions. Our experiments demonstrate that multi-objective methods that directly optimize the volume of the Pareto reward surface perform better and achieve a better balance of all rewards than those that attempt to find monotonic update directions. | [
{
"version": "v1",
"created": "Sun, 18 Feb 2024 21:25:09 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Oct 2024 21:51:03 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Jafari",
"Yasaman",
""
],
[
"Mekala",
"Dheeraj",
""
],
[
"Yu",
"Rose",
""
],
[
"Berg-Kirkpatrick",
"Taylor",
""
]
] |
2402.13184 | Mingyu Jin | Zhaoqian Xue, Beichen Wang, Suiyuan Zhu, Kai Mei, Hua Tang, Wenyue Hua, Mengnan Du, Yongfeng Zhang | What if LLMs Have Different World Views: Simulating Alien Civilizations with LLM-based Agents | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study introduces "CosmoAgent," an innovative artificial intelligence system that utilizes Large Language Models (LLMs) to simulate complex interactions between human and extraterrestrial civilizations. This paper introduces a mathematical model for quantifying the levels of civilization development and further employs a state transition matrix approach to evaluate their trajectories. Through this methodology, our study quantitatively analyzes the growth trajectories of civilizations, providing insights into future decision-making at critical points of growth and saturation. Furthermore, this paper acknowledges the vast diversity of potential living conditions across the universe, which could foster unique cosmologies, ethical codes, and worldviews among different civilizations. Recognizing the Earth-centric bias inherent in current LLM designs, we propose the novel concept of using LLM agents with diverse ethical paradigms and simulating interactions between entities with distinct moral principles. This innovative research not only introduces a novel method for comprehending potential inter-civilizational dynamics but also holds practical value in enabling entities with divergent value systems to strategize, prevent conflicts, and engage in games under conditions of asymmetric information. The accompanying code is available at https://github.com/MingyuJ666/Simulating-Alien-Civilizations-with-LLM-based-Agents. | [
{
"version": "v1",
"created": "Tue, 20 Feb 2024 17:49:46 GMT"
},
{
"version": "v2",
"created": "Wed, 21 Feb 2024 04:54:22 GMT"
},
{
"version": "v3",
"created": "Mon, 05 Aug 2024 04:45:55 GMT"
},
{
"version": "v4",
"created": "Fri, 20 Dec 2024 22:44:43 GMT"
},
{
"version": "v5",
"created": "Wed, 01 Jan 2025 23:34:53 GMT"
},
{
"version": "v6",
"created": "Mon, 09 Jun 2025 00:57:35 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Xue",
"Zhaoqian",
""
],
[
"Wang",
"Beichen",
""
],
[
"Zhu",
"Suiyuan",
""
],
[
"Mei",
"Kai",
""
],
[
"Tang",
"Hua",
""
],
[
"Hua",
"Wenyue",
""
],
[
"Du",
"Mengnan",
""
],
[
"Zhang",
"Yongfeng",
""
]
] |
2402.13217 | Long Zhao | Long Zhao, Nitesh B. Gundavarapu, Liangzhe Yuan, Hao Zhou, Shen Yan, Jennifer J. Sun, Luke Friedman, Rui Qian, Tobias Weyand, Yue Zhao, Rachel Hornung, Florian Schroff, Ming-Hsuan Yang, David A. Ross, Huisheng Wang, Hartwig Adam, Mikhail Sirotenko, Ting Liu, Boqing Gong | VideoPrism: A Foundational Visual Encoder for Video Understanding | Accepted to ICML 2024. v2: added retrieval results on MSRVTT (1K-A), more data analyses, and ablation studies; v3: released models at https://github.com/google-deepmind/videoprism | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | We introduce VideoPrism, a general-purpose video encoder that tackles diverse video understanding tasks with a single frozen model. We pretrain VideoPrism on a heterogeneous corpus containing 36M high-quality video-caption pairs and 582M video clips with noisy parallel text (e.g., ASR transcripts). The pretraining approach improves upon masked autoencoding by global-local distillation of semantic video embeddings and a token shuffling scheme, enabling VideoPrism to focus primarily on the video modality while leveraging the invaluable text associated with videos. We extensively test VideoPrism on four broad groups of video understanding tasks, from web video question answering to CV for science, achieving state-of-the-art performance on 31 out of 33 video understanding benchmarks. Our models are released at https://github.com/google-deepmind/videoprism. | [
{
"version": "v1",
"created": "Tue, 20 Feb 2024 18:29:49 GMT"
},
{
"version": "v2",
"created": "Sun, 16 Jun 2024 00:56:08 GMT"
},
{
"version": "v3",
"created": "Sat, 07 Jun 2025 01:16:44 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Zhao",
"Long",
""
],
[
"Gundavarapu",
"Nitesh B.",
""
],
[
"Yuan",
"Liangzhe",
""
],
[
"Zhou",
"Hao",
""
],
[
"Yan",
"Shen",
""
],
[
"Sun",
"Jennifer J.",
""
],
[
"Friedman",
"Luke",
""
],
[
"Qian",
"Rui",
""
],
[
"Weyand",
"Tobias",
""
],
[
"Zhao",
"Yue",
""
],
[
"Hornung",
"Rachel",
""
],
[
"Schroff",
"Florian",
""
],
[
"Yang",
"Ming-Hsuan",
""
],
[
"Ross",
"David A.",
""
],
[
"Wang",
"Huisheng",
""
],
[
"Adam",
"Hartwig",
""
],
[
"Sirotenko",
"Mikhail",
""
],
[
"Liu",
"Ting",
""
],
[
"Gong",
"Boqing",
""
]
] |
2402.14264 | Jikai Jin | Jikai Jin and Vasilis Syrgkanis | Structure-agnostic Optimality of Doubly Robust Learning for Treatment Effect Estimation | 31 pages, to appear in COLT 2025 | null | null | null | stat.ML cs.LG econ.EM math.ST stat.ME stat.TH | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Average treatment effect estimation is the most central problem in causal inference with application to numerous disciplines. While many estimation strategies have been proposed in the literature, the statistical optimality of these methods has still remained an open area of investigation, especially in regimes where these methods do not achieve parametric rates. In this paper, we adopt the recently introduced structure-agnostic framework of statistical lower bounds, which poses no structural properties on the nuisance functions other than access to black-box estimators that achieve some statistical estimation rate. This framework is particularly appealing when one is only willing to consider estimation strategies that use non-parametric regression and classification oracles as black-box sub-processes. Within this framework, we prove the statistical optimality of the celebrated and widely used doubly robust estimators for both the Average Treatment Effect (ATE) and the Average Treatment Effect on the Treated (ATT), as well as weighted variants of the former, which arise in policy evaluation. | [
{
"version": "v1",
"created": "Thu, 22 Feb 2024 04:03:32 GMT"
},
{
"version": "v2",
"created": "Sat, 02 Mar 2024 02:00:58 GMT"
},
{
"version": "v3",
"created": "Mon, 05 May 2025 01:30:17 GMT"
},
{
"version": "v4",
"created": "Sun, 08 Jun 2025 06:47:35 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Jin",
"Jikai",
""
],
[
"Syrgkanis",
"Vasilis",
""
]
] |
2402.15474 | Mohammed Osman Alawed | Mohammed Osman, Defu Hou, Wentao Wang, Hui Zhang | Functional renormalization group study of the quark-meson model with omega and rho vector mesons | null | null | null | null | hep-ph nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We employ the functional renormalization group flow equations to investigate the phase structure of the two-flavor quark-meson model in the presence of a finite isospin chemical potential, incorporating interactions with omega and rho vector mesons. For comparison, we also compute the phase diagram in the chiral limit using the mean-field approximation. Our findings demonstrate that omega and rho mesons affect the phase structure in markedly distinct ways, and the introduction of an isospin chemical potential leads to significant modifications in the phase boundaries and critical region. Increasing the isospin chemical potential lowers the tricritical points temperature, and tends to suppress the unphysical ``back-bending" of the FRG phase boundary at low temperature. | [
{
"version": "v1",
"created": "Fri, 23 Feb 2024 18:06:26 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Mar 2025 15:10:11 GMT"
},
{
"version": "v3",
"created": "Sun, 08 Jun 2025 13:14:33 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Osman",
"Mohammed",
""
],
[
"Hou",
"Defu",
""
],
[
"Wang",
"Wentao",
""
],
[
"Zhang",
"Hui",
""
]
] |
2402.15486 | Maria Carolina Bazotte Corgozinho | Maria Bazotte, Margarida Carvalho, and Thibaut Vidal | Solving Two-Stage Stochastic Programs with Endogenous Uncertainty via Random Variable Transformation | null | INFORMS Journal on Computing, 2025 | 10.1287/ijoc.2024.0847 | null | math.OC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Real-world decision-making problems often involve decision-dependent uncertainty, where the probability distribution of the random vector depends on the model decisions. Few studies focus on two-stage stochastic programs with this type of endogenous uncertainty, and those that do lack general methodologies. We propose a general method for solving a class of these programs based on random variable transformation, a technique widely employed in probability and statistics. The random variable transformation converts a stochastic program with endogenous uncertainty (original program) into an equivalent stochastic program with decision-independent uncertainty (transformed program), for which solution procedures are well-studied. Additionally, endogenous uncertainty usually leads to nonlinear nonconvex programs, which are theoretically intractable. Nonetheless, we show that, for some classical endogenous distributions, the proposed method yields mixed-integer linear or convex programs with exogenous uncertainty. We validate this method by applying it to a network design and facility-protection problem, considering distinct decision-dependent distributions for the random variables. While the original formulation of this problem is nonlinear nonconvex for most endogenous distributions, the proposed method transforms it into mixed-integer linear programs with exogenous uncertainty. We solve these transformed programs with the sample average approximation method. We highlight the superior performance of our approach compared to solving the original program in the case a mixed-integer linear formulation of this program exists. | [
{
"version": "v1",
"created": "Fri, 23 Feb 2024 18:26:16 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Jul 2024 03:14:13 GMT"
},
{
"version": "v3",
"created": "Thu, 18 Jul 2024 01:32:24 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Bazotte",
"Maria",
""
],
[
"Carvalho",
"Margarida",
""
],
[
"Vidal",
"Thibaut",
""
]
] |
2402.16887 | Jingtao Ding | Jingtao Ding, Chang Liu, Yu Zheng, Yunke Zhang, Zihan Yu, Ruikun Li, Hongyi Chen, Jinghua Piao, Huandong Wang, Jiazhen Liu and Yong Li | A Comprehensive Survey on Artificial Intelligence for Complex Network: Potential, Methodology and Application | 51 pages, 4 figures, 10 tables | null | null | null | cs.SI cs.AI cs.LG physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Complex networks pervade various real-world systems, from the natural environment to human societies. The essence of these networks is in their ability to transition and evolve from microscopic disorder-where network topology and node dynamics intertwine-to a macroscopic order characterized by certain collective behaviors. Over the past two decades, complex network science has significantly enhanced our understanding of the statistical mechanics, structures, and dynamics underlying real-world networks. Despite these advancements, there remain considerable challenges in exploring more realistic systems and enhancing practical applications. The emergence of artificial intelligence (AI) technologies, coupled with the abundance of diverse real-world network data, has heralded a new era in complex network science research. This survey aims to systematically address the potential advantages of AI in overcoming the lingering challenges of complex network research. It endeavors to summarize the pivotal research problems and provide an exhaustive review of the corresponding methodologies and applications. Through this comprehensive survey-the first of its kind on AI for complex networks-we expect to provide valuable insights that will drive further research and advancement in this interdisciplinary field. | [
{
"version": "v1",
"created": "Fri, 23 Feb 2024 09:06:36 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 01:40:44 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Ding",
"Jingtao",
""
],
[
"Liu",
"Chang",
""
],
[
"Zheng",
"Yu",
""
],
[
"Zhang",
"Yunke",
""
],
[
"Yu",
"Zihan",
""
],
[
"Li",
"Ruikun",
""
],
[
"Chen",
"Hongyi",
""
],
[
"Piao",
"Jinghua",
""
],
[
"Wang",
"Huandong",
""
],
[
"Liu",
"Jiazhen",
""
],
[
"Li",
"Yong",
""
]
] |
2402.17391 | Alexander Mikhalychev | Alexander Mikhalychev, Aleksandr Saushin, Alex Ulyanenkov, and Polina Kuzhir | Spatial super-resolution in nanosensing with blinking emitters | null | null | null | null | physics.optics quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a method of spatial resolution enhancement in metrology (thermometry, magnetometry, pH estimation, and similar methods) with blinking fluorescent nanosensors by combining sensing with super-resolution optical fluctuation imaging (SOFI). To demonstrate efficiency of this approach, both numerical simulations and a model experiment with laser diodes, which model fluctuating nanoemitters, and intentional blurring of the image are performed. The 2nd, 3rd, and 4th order cumulant images provide improvement of the contrast and enable successful reconstruction of smaller features of the modeled temperature (or any other physical parameter) distribution relatively to the intensity-based approach. We believe that blinking fluorescent sensing agents being complemented with the developed image analysis technique could be utilized routinely in the life science sector for recognizing the local changes in the spectral response of blinking fluorophores, e.g. delivered targetly to the wanted cell or even organelle. It is extremely useful for the local measurements of living cells' physical parameters changes due to applying any external "forces", including disease effect, aging, healing, or response to the treatment. | [
{
"version": "v1",
"created": "Tue, 27 Feb 2024 10:38:05 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 14:33:15 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Mikhalychev",
"Alexander",
""
],
[
"Saushin",
"Aleksandr",
""
],
[
"Ulyanenkov",
"Alex",
""
],
[
"Kuzhir",
"Polina",
""
]
] |
2402.17481 | Steven Walton | Steven Walton and Minh-Binh Tran | Numerical Schemes for 3-Wave Kinetic Equations: A Complete Treatment of the Collision Operator | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In our previous work, numerical schemes for a simplified version of 3-wave kinetic equations, in which only the simple forward-cascade terms of the collision operators are kept, have been successfully designed, especially to capture the long time dynamics of the equation given the multiple blow-up time phenomenon. In this second work in the series, we propose numerical treatments for the complete 3-wave kinetic equations, in which the complete, much more complicated collision operators are fully considered based on a novel conservative form of the equation. We then derive an implicit finite volume scheme to solve the equation. The new discretization uses an adaptive time-stepping method which allows for the simulations to be carried to very long times. Our computed solutions are compared with previously derived long-time asymptotic estimates for the decay rate of total energy of time-dependent solutions of 3-wave kinetic equations and found to be in excellent agreement. | [
{
"version": "v1",
"created": "Tue, 27 Feb 2024 13:08:26 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 00:46:30 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Walton",
"Steven",
""
],
[
"Tran",
"Minh-Binh",
""
]
] |
2402.18893 | Tien-Ming Chuang | Balaji Venkatesan, Syu-You Guan, Jen-Te Chang, Shiang-Bin Chiu, Po-Yuan Yang, Chih-Chuan Su, Tay-Rong Chang, Kalaivanan Raju, Raman Sankar, Somboon Fongchaiya, Ming-Wen Chu, Chia-Seng Chang, Guoqing Chang, Hsin Lin, Adrian Del Maestro, Ying-Jer Kao and Tien-Ming Chuang | Direct Visualization of a Disorder Driven Electronic Smectic Phase in Dirac Nodal Line Semimetal GdSbTe | null | npj Quantum Mater. 10, 56 (2025) | 10.1038/s41535-025-00779-y | null | cond-mat.str-el cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Electronic liquid crystal (ELC) phases are spontaneous symmetry breaking states believed to arise from strong electron correlation in quantum materials such as cuprates and iron pnictides. Here, we report a direct observation of a smectic phase in a weakly correlated Dirac nodal line (DNL) semimetal GdSbxTe2-x. Incommensurate smectic charge modulation and intense local unidirectional nanostructure are visualized by using spectroscopic imaging - scanning tunneling microscopy. As topological materials with symmetry protected Dirac or Weyl fermions are mostly weakly correlated, the discovery of such an ELC phase are anomalous and raise questions on the origin of their emergence. Specifically, we demonstrate how chemical substitution generates these symmetry breaking phases before the system undergoes a charge density wave (CDW) - orthorhombic structural transition. Our results highlight the importance of impurities in realizing ELC phases and present a new material platform for exploring the interplay among quenched disorder, topology and electron correlation. | [
{
"version": "v1",
"created": "Thu, 29 Feb 2024 06:31:57 GMT"
},
{
"version": "v2",
"created": "Tue, 07 May 2024 18:14:07 GMT"
},
{
"version": "v3",
"created": "Thu, 02 Jan 2025 06:10:35 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Venkatesan",
"Balaji",
""
],
[
"Guan",
"Syu-You",
""
],
[
"Chang",
"Jen-Te",
""
],
[
"Chiu",
"Shiang-Bin",
""
],
[
"Yang",
"Po-Yuan",
""
],
[
"Su",
"Chih-Chuan",
""
],
[
"Chang",
"Tay-Rong",
""
],
[
"Raju",
"Kalaivanan",
""
],
[
"Sankar",
"Raman",
""
],
[
"Fongchaiya",
"Somboon",
""
],
[
"Chu",
"Ming-Wen",
""
],
[
"Chang",
"Chia-Seng",
""
],
[
"Chang",
"Guoqing",
""
],
[
"Lin",
"Hsin",
""
],
[
"Del Maestro",
"Adrian",
""
],
[
"Kao",
"Ying-Jer",
""
],
[
"Chuang",
"Tien-Ming",
""
]
] |
2402.19331 | X. Griffin Wang | X. Griffin Wang | Multiplicative Hitchin Fibrations and the Fundamental Lemma | 265 pages. Accepted, to appear as an Annals of Math Studies volume. Revision addressing referees' comments: (1) substantially restructure and rewrite part of chapter 5, especially the current 5.2; (2) upgrade the product formula in 6.9 using results from newer literature; (3) numerous small fixes and adjustments suggested by the referees; (4) add an index and list of notations | null | null | null | math.AG math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let $k$ be a finite field and let $G$ be a reductive group over $k[[\pi]]$. Suppose $\mathrm{char}(k)$ is larger than twice the Coxeter number of $G$, we prove the standard endoscopic fundamental lemma for the spherical Hecke algebra of $G$ using multiplicative Hitchin fibrations. | [
{
"version": "v1",
"created": "Thu, 29 Feb 2024 16:33:35 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Oct 2024 00:53:28 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 14:22:17 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Wang",
"X. Griffin",
""
]
] |
2403.00909 | Luca Di Mascolo | Luca Di Mascolo, Yvette Perrott, Tony Mroczkowski, Srinivasan Raghunathan, Stefano Andreon, Stefano Ettori, Aurora Simionescu, Joshiwa van Marrewijk, Claudia Cicone, Minju Lee, Dylan Nelson, Laura Sommovigo, Mark Booth, Pamela Klaassen, Paola Andreani, Martin A. Cordiner, Doug Johnstone, Eelco van Kampen, Daizhong Liu, Thomas J. Maccarone, Thomas W. Morris, John Orlowski-Scherer, Am\'elie Saintonge, Matthew Smith, Alexander E. Thelen, and Sven Wedemeyer | Atacama Large Aperture Submillimeter Telescope (AtLAST) Science: Resolving the Hot and Ionized Universe through the Sunyaev-Zeldovich effect | 39 pages, 18 figures, 1 table. Published in Open Research Europe as part of the AtLAST Design Study collection: see https://open-research-europe.ec.europa.eu/collections/atlast/about | Open Research Europe, 4:113 (2025) | 10.12688/openreseurope.17449.2 | null | astro-ph.CO astro-ph.GA astro-ph.IM | http://creativecommons.org/licenses/by/4.0/ | An omnipresent feature of the multi-phase ``cosmic web'' is that warm/hot (>$10^5$ K) ionized gas pervades it. This gas constitutes a relevant contribution to the overall universal matter budget across multiple scales, from the several tens of Mpc-scale IGM filaments, to the Mpc ICM, all the way down to the CGM surrounding individual galaxies from ~1 kpc up to their respective virial radii (~100 kpc). The study of the hot baryonic component of cosmic matter density represents a powerful means for constraining the intertwined evolution of galactic populations and large-scale cosmological structures, for tracing the matter assembly in the Universe and its thermal history. To this end, the SZ effect provides the ideal observational tool for measurements out to the beginnings of structure formation. The SZ effect is caused by the scattering of the photons from the cosmic microwave background off the hot electrons embedded within cosmic structures, and provides a redshift-independent perspective on the thermal and kinematic properties of the warm/hot gas. Still, current and future (sub)mm facilities have been providing only a partial view of the SZ Universe due to any combination of: limited angular resolution, spectral coverage, field of view, spatial dynamic range, sensitivity. In this paper, we motivate the development of a wide-field, broad-band, multi-chroic continuum instrument for the Atacama Large Aperture Submillimeter Telescope (AtLAST) by identifying the scientific drivers that will deepen our understanding of the complex thermal evolution of cosmic structures. On a technical side, this will necessarily require efficient multi-wavelength mapping of the SZ signal with an unprecedented spatial dynamic range (from arcsecond to tens of arcminutes) and we employ theoretical forecasts to determine the key instrumental constraints for achieving our goals. [abridged] | [
{
"version": "v1",
"created": "Fri, 01 Mar 2024 19:00:02 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 17:13:27 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Di Mascolo",
"Luca",
""
],
[
"Perrott",
"Yvette",
""
],
[
"Mroczkowski",
"Tony",
""
],
[
"Raghunathan",
"Srinivasan",
""
],
[
"Andreon",
"Stefano",
""
],
[
"Ettori",
"Stefano",
""
],
[
"Simionescu",
"Aurora",
""
],
[
"van Marrewijk",
"Joshiwa",
""
],
[
"Cicone",
"Claudia",
""
],
[
"Lee",
"Minju",
""
],
[
"Nelson",
"Dylan",
""
],
[
"Sommovigo",
"Laura",
""
],
[
"Booth",
"Mark",
""
],
[
"Klaassen",
"Pamela",
""
],
[
"Andreani",
"Paola",
""
],
[
"Cordiner",
"Martin A.",
""
],
[
"Johnstone",
"Doug",
""
],
[
"van Kampen",
"Eelco",
""
],
[
"Liu",
"Daizhong",
""
],
[
"Maccarone",
"Thomas J.",
""
],
[
"Morris",
"Thomas W.",
""
],
[
"Orlowski-Scherer",
"John",
""
],
[
"Saintonge",
"Amélie",
""
],
[
"Smith",
"Matthew",
""
],
[
"Thelen",
"Alexander E.",
""
],
[
"Wedemeyer",
"Sven",
""
]
] |
2403.02036 | Junjie Zeng | Junjie Zeng, James Jun He, Zhen Ning, Dong-Hui Xu, and Rui Wang | Spin Signature of Majorana Fermions in Topological Nodal-Point Superconductors | 6 pages, 3 figures | npj Quantum Mater. 10, 48 (2025) | 10.1038/s41535-025-00768-1 | null | cond-mat.supr-con cond-mat.mes-hall | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In two-dimensional topological nodal superconductors, Majorana edge states have been conventionally believed to exhibit only spin-triplet pairing correlations. However, we reveal a substantial spin-singlet pairing component in Majorana edge states of antiferromagnetic topological nodal-point superconductors. This unexpected phenomenon emerges from the interplay between antiferromagnetic order and symmetry, resulting in Majorana edge states with a nearly flat band dispersion, deviating from the strictly flat band. Crucially, this phenomenon is detectable through spin-selective Andreev reflection, where the zero-bias conductance peaks are maximized when the spin of incident electrons is nearly antiparallel to that of Majorana edge excitations. This discovery unveils a unique spin signature for Andreev reflection resonances, advancing our fundamental understanding of spin-dependent mechanisms in topological superconductivity and representing a significant step towards the experimental detection of Majorana fermions. | [
{
"version": "v1",
"created": "Mon, 04 Mar 2024 13:41:54 GMT"
},
{
"version": "v2",
"created": "Wed, 05 Jun 2024 07:07:53 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Zeng",
"Junjie",
""
],
[
"He",
"James Jun",
""
],
[
"Ning",
"Zhen",
""
],
[
"Xu",
"Dong-Hui",
""
],
[
"Wang",
"Rui",
""
]
] |
2403.04310 | Qinghua Lei | Qinghua Lei, Didier Sornette | Endo-exo framework for a unifying classification of episodic landslide movements | 7 figures | null | null | null | physics.geo-ph | http://creativecommons.org/licenses/by/4.0/ | Landslides exhibit intermittent gravity-driven downslope movements developing over days to years before a possible major collapse, commonly boosted by external events like precipitations and earthquakes. The reasons behind these episodic movements and how they relate to the final instability remain poorly understood. Here, we develop a novel "endo-exo" theory to quantitatively diagnose landslide dynamics, capturing the interplay between exogenous stressors such as rainfall and endogenous damage/healing processes. We predict four distinct types of episodic landslide dynamics (endogenous/exogenous-subcritical/critical), characterized by power law relaxations with different exponents, all related to a single parameter \vartheta. These predictions are tested on the dataset of the Preonzo landslide, which exhibited multi-year episodic movements prior to a catastrophic collapse. All its sporadic activities can be accounted for within this classification with \vartheta \approx 0.45\pm0.1, providing strong support for our parsimonious theory. We find that the final collapse of this landslide is clearly preceded over 1-2 months by an increased frequency of medium/large velocities, signaling the transition into a catastrophic regime with amplifying positive feedbacks. Our research suggests that landslides may not permanently operate at a critical state, which has major implications for forecasting catastrophic failure events. | [
{
"version": "v1",
"created": "Thu, 07 Mar 2024 08:30:18 GMT"
},
{
"version": "v2",
"created": "Sun, 08 Jun 2025 12:12:00 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Lei",
"Qinghua",
""
],
[
"Sornette",
"Didier",
""
]
] |
2403.04620 | Vladislav Vysotsky | Vladislav Vysotsky | Stationary switching random walks | Effectively, this is a new paper. We added: 1) New uniqueness results, which comprise the main contribution of the paper. 2) The criteria for the topological recurrence. 3) Proof of topological irreducibility. 4) Applications to reflected random walks. 5) Applications to GI/G/1 queues with vacation. 6) An extended overview of the literature | null | null | null | math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A switching random walk, commonly known under the misnomer `oscillating random walk', is a real-valued Markov chain whose distribution of increments is determined by the sign of the current position. We explicitly identify an invariant measure of this chain and study its uniqueness, up to a constant factor, within the class of locally finite invariant measures. Next we provide sufficient conditions for the topological recurrence of the switching random walk, and prove its topological irreducibility on a suitably chosen state space. As a consequence of our approach, we establish a new connection between the classical stationary distributions of the renewal theory and stationarity of the Lebesgue measure for random walks. We give further applications concerning reflected random walks and the waiting times in GI/G/1 queues with vacation. | [
{
"version": "v1",
"created": "Thu, 07 Mar 2024 16:03:13 GMT"
},
{
"version": "v2",
"created": "Sun, 08 Jun 2025 21:50:16 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Vysotsky",
"Vladislav",
""
]
] |
2403.04725 | Daniele De Gennaro | Antonin Chambolle, Daniele De Gennaro, Massimiliano Morini | Discrete-to-continuum crystalline curvature flows | null | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | We consider here a fully discrete variant of the implicit variational scheme for mean curvature flow [AlmTayWan,LucStu], in a setting where the flow is governed by a crystalline surface tension defined by the limit of pairwise interactions energy on the discrete grid. The algorithm is based on a new discrete distance from the evolving sets, which prevents the occurrence of the spatial drift and pinning phenomena identified in [MisiatsYip16,BraGelNov] in a similar discrete framework. We provide the first rigorous convergence result holding in any dimension, for any initial set and for a large class of purely crystalline anisotropies, in which the spatial discretization mesh can be of the same order or coarser than the time step. | [
{
"version": "v1",
"created": "Thu, 07 Mar 2024 18:23:45 GMT"
},
{
"version": "v2",
"created": "Sat, 30 Mar 2024 09:51:40 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 10:38:45 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Chambolle",
"Antonin",
""
],
[
"De Gennaro",
"Daniele",
""
],
[
"Morini",
"Massimiliano",
""
]
] |
2403.04764 | Zhaolin Ren | Zhaolin Ren and Na Li | TS-RSR: A provably efficient approach for batch Bayesian Optimization | Accepted by the SIAM Journal on Optimization | null | null | null | cs.LG math.OC stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a new approach for batch Bayesian Optimization (BO) called Thompson Sampling-Regret to Sigma Ratio directed sampling (TS-RSR), where we sample a new batch of actions by minimizing a Thompson Sampling approximation of a regret to uncertainty ratio. Our sampling objective is able to coordinate the actions chosen in each batch in a way that minimizes redundancy between points whilst focusing on points with high predictive means or high uncertainty. Theoretically, we provide rigorous convergence guarantees on our algorithm's regret, and numerically, we demonstrate that our method attains state-of-the-art performance on a range of challenging synthetic and realistic test functions, where it outperforms several competitive benchmark batch BO algorithms. | [
{
"version": "v1",
"created": "Thu, 07 Mar 2024 18:58:26 GMT"
},
{
"version": "v2",
"created": "Tue, 09 Apr 2024 01:28:43 GMT"
},
{
"version": "v3",
"created": "Thu, 02 May 2024 14:16:35 GMT"
},
{
"version": "v4",
"created": "Mon, 09 Jun 2025 15:42:30 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Ren",
"Zhaolin",
""
],
[
"Li",
"Na",
""
]
] |
2403.05369 | Linwei Chen | Linwei Chen, Lin Gu, Ying Fu | Frequency-Adaptive Dilated Convolution for Semantic Segmentation | CVPR 2024 highlighted | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dilated convolution, which expands the receptive field by inserting gaps between its consecutive elements, is widely employed in computer vision. In this study, we propose three strategies to improve individual phases of dilated convolution from the view of spectrum analysis. Departing from the conventional practice of fixing a global dilation rate as a hyperparameter, we introduce Frequency-Adaptive Dilated Convolution (FADC), which dynamically adjusts dilation rates spatially based on local frequency components. Subsequently, we design two plug-in modules to directly enhance effective bandwidth and receptive field size. The Adaptive Kernel (AdaKern) module decomposes convolution weights into low-frequency and high-frequency components, dynamically adjusting the ratio between these components on a per-channel basis. By increasing the high-frequency part of convolution weights, AdaKern captures more high-frequency components, thereby improving effective bandwidth. The Frequency Selection (FreqSelect) module optimally balances high- and low-frequency components in feature representations through spatially variant reweighting. It suppresses high frequencies in the background to encourage FADC to learn a larger dilation, thereby increasing the receptive field for an expanded scope. Extensive experiments on segmentation and object detection consistently validate the efficacy of our approach. The code is publicly available at https://github.com/Linwei-Chen/FADC. | [
{
"version": "v1",
"created": "Fri, 08 Mar 2024 15:00:44 GMT"
},
{
"version": "v2",
"created": "Tue, 12 Mar 2024 08:33:51 GMT"
},
{
"version": "v3",
"created": "Mon, 18 Mar 2024 01:53:42 GMT"
},
{
"version": "v4",
"created": "Thu, 28 Mar 2024 13:41:05 GMT"
},
{
"version": "v5",
"created": "Mon, 01 Apr 2024 07:26:06 GMT"
},
{
"version": "v6",
"created": "Tue, 21 May 2024 14:29:31 GMT"
},
{
"version": "v7",
"created": "Mon, 09 Jun 2025 03:30:05 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Chen",
"Linwei",
""
],
[
"Gu",
"Lin",
""
],
[
"Fu",
"Ying",
""
]
] |
2403.06054 | Xiang Li | Xiang Li, Soo Min Kwon, Shijun Liang, Ismail R. Alkhouri, Saiprasad Ravishankar, Qing Qu | Decoupled Data Consistency with Diffusion Purification for Image Restoration | null | null | null | null | eess.IV cs.AI cs.CV cs.LG eess.SP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Diffusion models have recently gained traction as a powerful class of deep generative priors, excelling in a wide range of image restoration tasks due to their exceptional ability to model data distributions. To solve image restoration problems, many existing techniques achieve data consistency by incorporating additional likelihood gradient steps into the reverse sampling process of diffusion models. However, the additional gradient steps pose a challenge for real-world practical applications as they incur a large computational overhead, thereby increasing inference time. They also present additional difficulties when using accelerated diffusion model samplers, as the number of data consistency steps is limited by the number of reverse sampling steps. In this work, we propose a novel diffusion-based image restoration solver that addresses these issues by decoupling the reverse process from the data consistency steps. Our method involves alternating between a reconstruction phase to maintain data consistency and a refinement phase that enforces the prior via diffusion purification. Our approach demonstrates versatility, making it highly adaptable for efficient problem-solving in latent space. Additionally, it reduces the necessity for numerous sampling steps through the integration of consistency models. The efficacy of our approach is validated through comprehensive experiments across various image restoration tasks, including image denoising, deblurring, inpainting, and super-resolution. | [
{
"version": "v1",
"created": "Sun, 10 Mar 2024 00:47:05 GMT"
},
{
"version": "v2",
"created": "Tue, 12 Mar 2024 03:22:13 GMT"
},
{
"version": "v3",
"created": "Fri, 22 Mar 2024 19:14:59 GMT"
},
{
"version": "v4",
"created": "Wed, 27 Mar 2024 17:06:10 GMT"
},
{
"version": "v5",
"created": "Wed, 29 May 2024 00:09:08 GMT"
},
{
"version": "v6",
"created": "Sun, 08 Jun 2025 16:36:27 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Li",
"Xiang",
""
],
[
"Kwon",
"Soo Min",
""
],
[
"Liang",
"Shijun",
""
],
[
"Alkhouri",
"Ismail R.",
""
],
[
"Ravishankar",
"Saiprasad",
""
],
[
"Qu",
"Qing",
""
]
] |
2403.07061 | Elizabeth Bennewitz | Elizabeth R. Bennewitz, Brayden Ware, Alexander Schuckert, Alessio Lerose, Federica M. Surace, Ron Belyansky, William Morong, De Luo, Arinjoy De, Kate S. Collins, Or Katz, Christopher Monroe, Zohreh Davoudi, Alexey V. Gorshkov | Simulating Meson Scattering on Spin Quantum Simulators | 18 pages, 4 main figures, 2 supplementary figures | null | null | null | quant-ph cond-mat.quant-gas hep-ph nucl-th | http://creativecommons.org/licenses/by/4.0/ | Studying high-energy collisions of composite particles, such as hadrons and nuclei, is an outstanding goal for quantum simulators. However, preparation of hadronic wave packets has posed a significant challenge, due to the complexity of hadrons and the precise structure of wave packets. This has limited demonstrations of hadron scattering on quantum simulators to date. Observations of confinement and composite excitations in quantum spin systems have opened up the possibility to explore scattering dynamics in spin models. In this article, we develop two methods to create entangled spin states corresponding to wave packets of composite particles in analog quantum simulators of Ising spin Hamiltonians. One wave-packet preparation method uses the blockade effect enabled by beyond-nearest-neighbor Ising spin interactions. The other method utilizes a quantum-bus-mediated exchange, such as the native spin-phonon coupling in trapped-ion arrays. With a focus on trapped-ion simulators, we numerically benchmark both methods and show that high-fidelity wave packets can be achieved in near-term experiments. We numerically study scattering of wave packets for experimentally realizable parameters in the Ising model and find inelastic-scattering regimes, corresponding to particle production in the scattering event, with prominent and distinct experimental signals. Our proposal, therefore, demonstrates the potential of observing inelastic scattering in near-term quantum simulators. | [
{
"version": "v1",
"created": "Mon, 11 Mar 2024 18:00:07 GMT"
},
{
"version": "v2",
"created": "Wed, 13 Mar 2024 14:46:25 GMT"
},
{
"version": "v3",
"created": "Sun, 08 Jun 2025 21:22:49 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Bennewitz",
"Elizabeth R.",
""
],
[
"Ware",
"Brayden",
""
],
[
"Schuckert",
"Alexander",
""
],
[
"Lerose",
"Alessio",
""
],
[
"Surace",
"Federica M.",
""
],
[
"Belyansky",
"Ron",
""
],
[
"Morong",
"William",
""
],
[
"Luo",
"De",
""
],
[
"De",
"Arinjoy",
""
],
[
"Collins",
"Kate S.",
""
],
[
"Katz",
"Or",
""
],
[
"Monroe",
"Christopher",
""
],
[
"Davoudi",
"Zohreh",
""
],
[
"Gorshkov",
"Alexey V.",
""
]
] |
2403.07931 | Junyu Liu | Junyu Liu, Wangkai Jin, Xiangjun Peng | Formalizing Feint Actions, and Example Studies in Two-Player Games | The pre-print has been combined into another paper (arXiv:2403.07932) | null | null | null | cs.GT cs.GR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Feint actions refer to a set of deceptive actions, which enable players to obtain temporal advantages from their opponents. Such actions are regarded as widely-used tactic in most non-deterministic Two-player Games (e.g. boxing and fencing). However, existing literature does not provide comprehensive and concrete formalization on Feint actions, and their implications on Two-Player Games. We argue that a full exploration on Feint actions is of great importance towards more realistic Two-player Games. In this paper, we provide the first comprehensive and concrete formalization of Feint actions. The key idea of our work is to (1) allow automatic generation of Feint actions, via our proposed Palindrome-directed Generation of Feint actions; and (2) provide concrete principles to properly combine Feint and attack actions. Based on our formalization of Feint actions, we also explore the implications on the game strategy model, and provide optimizations to better incorporate Feint actions. Our experimental results shows that accounting for Feint actions in Non-Deterministic Games (1) brings overall benefits to the game design; and (2) has great benefits on on either game animations or strategy designs, which also introduces a great extent of randomness into randomness-demanded Game models. | [
{
"version": "v1",
"created": "Sun, 03 Mar 2024 18:23:18 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 16:23:45 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Liu",
"Junyu",
""
],
[
"Jin",
"Wangkai",
""
],
[
"Peng",
"Xiangjun",
""
]
] |
2403.07932 | Junyu Liu | Junyu Liu, Xiangjun Peng | Feint Behaviors and Strategies: Formalization, Implementation and Evaluation | Published in NeurIPS 2024 | null | null | null | cs.GT cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Feint behaviors refer to a set of deceptive behaviors in a nuanced manner, which enable players to obtain temporal and spatial advantages over opponents in competitive games. Such behaviors are crucial tactics in most competitive multi-player games (e.g., boxing, fencing, basketball, motor racing, etc.). However, existing literature does not provide a comprehensive (and/or concrete) formalization for Feint behaviors, and their implications on game strategies. In this work, we introduce the first comprehensive formalization of Feint behaviors at both action-level and strategy-level, and provide concrete implementation and quantitative evaluation of them in multi-player games. The key idea of our work is to (1) allow automatic generation of Feint behaviors via Palindrome-directed templates, combine them into meaningful behavior sequences via a Dual-Behavior Model; (2) concertize the implications from our formalization of Feint on game strategies, in terms of temporal, spatial and their collective impacts respectively; and (3) provide a unified implementation scheme of Feint behaviors in existing MARL frameworks. The experimental results show that our design of Feint behaviors can (1) greatly improve the game reward gains; (2) significantly improve the diversity of Multi-Player Games; and (3) only incur negligible overheads in terms of time consumption. | [
{
"version": "v1",
"created": "Mon, 04 Mar 2024 03:43:45 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 16:22:33 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Liu",
"Junyu",
""
],
[
"Peng",
"Xiangjun",
""
]
] |
2403.11479 | Ruixuan Zhu | Yang Zhou and Ruixuan Zhu | A priori estimates for parabolic Monge-Amp\`ere type equations | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove the existence and regularity of convex solutions to the first initial-boundary value problem for the parabolic Monge-Amp\`ere equationn $$
\left\{\begin{eqnarray}
&&-u_t+\det D^2u= \psi(x,t) \quad\quad\ \text{ in } Q_T,\newline
&&u=\phi\quad\text{ on }\partial_pQ_T,
\end{eqnarray}\right. $$ where $\psi,\phi$ are given functions, $Q_T=\Omega\times(0,T]$, $\partial_p Q_T$ is the parabolic boundary of $Q_T$, and $\Omega\subset\mathbb{R}^n$ is a uniformly convex domain. Our approach can also be used to prove similar results for the $\gamma$-Gauss curvature flow with any $0<\gamma\le 1$. | [
{
"version": "v1",
"created": "Mon, 18 Mar 2024 04:55:00 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 09:31:49 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Zhou",
"Yang",
""
],
[
"Zhu",
"Ruixuan",
""
]
] |
2403.12334 | Bala Krishnamoorthy | Kostiantyn Lyman, Rounak Meyur, Bala Krishnamoorthy, Mahantesh Halappanavar | Structural Validation Of Synthetic Power Distribution Networks Using The Multiscale Flat Norm | Minor revision; to appear in IJCGA | null | null | null | cs.CG math.AT math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the problem of comparing a pair of geometric networks that may not be similarly defined, i.e., when they do not have one-to-one correspondences between their nodes and edges. Our motivating application is to compare power distribution networks of a region. Due to the lack of openly available power network datasets, researchers synthesize realistic networks resembling their actual counterparts. But the synthetic digital twins may vary significantly from one another and from actual networks due to varying underlying assumptions and approaches. Hence the user wants to evaluate the quality of networks in terms of their structural similarity to actual power networks. But the lack of correspondence between the networks renders most standard approaches, e.g., subgraph isomorphism and edit distance, unsuitable.
We propose an approach based on the multiscale flat norm, a notion of distance between objects defined in the field of geometric measure theory, to compute the distance between a pair of planar geometric networks. Using a triangulation of the domain containing the input networks, the flat norm distance between two networks at a given scale can be computed by solving a linear program. In addition, this computation automatically identifies the 2D regions (patches) that capture where the two networks are different. We demonstrate through 2D examples that the flat norm distance can capture the variations of inputs more accurately than the commonly used Hausdorff distance. As a notion of stability, we also derive upper bounds on the flat norm distance between a simple 1D curve and its perturbed version as a function of the radius of perturbation for a restricted class of perturbations. We demonstrate our approach on a set of actual power networks from a county in the USA. | [
{
"version": "v1",
"created": "Tue, 19 Mar 2024 00:28:27 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 06:25:46 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Lyman",
"Kostiantyn",
""
],
[
"Meyur",
"Rounak",
""
],
[
"Krishnamoorthy",
"Bala",
""
],
[
"Halappanavar",
"Mahantesh",
""
]
] |
2403.12886 | Federico Nocentini | Federico Nocentini, Claudio Ferrari, Stefano Berretti | EmoVOCA: Speech-Driven Emotional 3D Talking Heads | WACV 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The domain of 3D talking head generation has witnessed significant progress in recent years. A notable challenge in this field consists in blending speech-related motions with expression dynamics, which is primarily caused by the lack of comprehensive 3D datasets that combine diversity in spoken sentences with a variety of facial expressions. Whereas literature works attempted to exploit 2D video data and parametric 3D models as a workaround, these still show limitations when jointly modeling the two motions. In this work, we address this problem from a different perspective, and propose an innovative data-driven technique that we used for creating a synthetic dataset, called EmoVOCA, obtained by combining a collection of inexpressive 3D talking heads and a set of 3D expressive sequences. To demonstrate the advantages of this approach, and the quality of the dataset, we then designed and trained an emotional 3D talking head generator that accepts a 3D face, an audio file, an emotion label, and an intensity value as inputs, and learns to animate the audio-synchronized lip movements with expressive traits of the face. Comprehensive experiments, both quantitative and qualitative, using our data and generator evidence superior ability in synthesizing convincing animations, when compared with the best performing methods in the literature. Our code and pre-trained model will be made available. | [
{
"version": "v1",
"created": "Tue, 19 Mar 2024 16:33:26 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Sep 2024 14:28:08 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 11:06:34 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Nocentini",
"Federico",
""
],
[
"Ferrari",
"Claudio",
""
],
[
"Berretti",
"Stefano",
""
]
] |
2403.12923 | Margarida Carvalho | Quang Minh Bui, Margarida Carvalho, Jos\'e Neto | Solving Combinatorial Pricing Problems using Embedded Dynamic Programming Models | null | INFORMS Journal on Computing, 2025 | 10.1287/ijoc.2024.0686 | null | math.OC cs.DM | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The combinatorial pricing problem (CPP) is a bilevel problem in which the leader maximizes their revenue by imposing tolls on certain items that they can control. Based on the tolls set by the leader, the follower selects a subset of items corresponding to an optimal solution of a combinatorial optimization problem. To accomplish the leader's goal, the tolls need to be sufficiently low to discourage the follower from choosing the items offered by the competitors. In this paper, we derive a single-level reformulation for the CPP by rewriting the follower's problem as a longest path problem using a dynamic programming model, and then taking its dual and applying strong duality. We proceed to solve the reformulation in a dynamic fashion with a cutting plane method. We apply this methodology to two distinct dynamic programming models, namely, a novel formulation designated as selection diagram and the well-known decision diagram. We also produce numerical results to evaluate their performances across three different specializations of the CPP and a closely related problem that is the knapsack interdiction problem. Our results showcase the potential of the two proposed reformulations over the natural value function approach, expanding the set of tools to solve combinatorial bilevel programs. | [
{
"version": "v1",
"created": "Tue, 19 Mar 2024 17:28:48 GMT"
},
{
"version": "v2",
"created": "Sat, 29 Mar 2025 09:11:11 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Bui",
"Quang Minh",
""
],
[
"Carvalho",
"Margarida",
""
],
[
"Neto",
"José",
""
]
] |
2403.13778 | Saeed Saadatnejad | Mohammadhossein Bahari, Saeed Saadatnejad, Amirhossein Askari Farsangi, Seyed-Mohsen Moosavi-Dezfooli, Alexandre Alahi | Certified Human Trajectory Prediction | CVPR 2025 | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Predicting human trajectories is essential for the safe operation of autonomous vehicles, yet current data-driven models often lack robustness in case of noisy inputs such as adversarial examples or imperfect observations. Although some trajectory prediction methods have been developed to provide empirical robustness, these methods are heuristic and do not offer guaranteed robustness. In this work, we propose a certification approach tailored for trajectory prediction that provides guaranteed robustness. To this end, we address the unique challenges associated with trajectory prediction, such as unbounded outputs and multi-modality. To mitigate the inherent performance drop through certification, we propose a diffusion-based trajectory denoiser and integrate it into our method. Moreover, we introduce new certified performance metrics to reliably measure the trajectory prediction performance. Through comprehensive experiments, we demonstrate the accuracy and robustness of the certified predictors and highlight their advantages over the non-certified ones. The code is available online: https://s-attack.github.io/. | [
{
"version": "v1",
"created": "Wed, 20 Mar 2024 17:41:35 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Jun 2025 18:34:56 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Bahari",
"Mohammadhossein",
""
],
[
"Saadatnejad",
"Saeed",
""
],
[
"Farsangi",
"Amirhossein Askari",
""
],
[
"Moosavi-Dezfooli",
"Seyed-Mohsen",
""
],
[
"Alahi",
"Alexandre",
""
]
] |
2403.14322 | Francesco Macheda | Francesco Macheda, Francesco Mauri and Thibault Sohier | Ab-initio Van der Waals electrodynamics: polaritons and electron scattering from plasmons and phonons in BN-capped graphene | 32 pages, 18 figures. Reuploaded with CC BY licence | Phys. Rev. B 110, 115407 (2024) | 10.1103/PhysRevB.110.115407 | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | Plasmons and polar phonons are elementary electrodynamic excitations of matter. In 2d and at long wavelengths, they couple to light and act as the system polaritons. They also dictate the scattering of charged carriers. Van der Waals heterostructures offer the opportunity to couple excitations from different layers via long-range Coulomb interactions, modifying both their dispersion and their scattering of electrons. Even when the excitations do not couple, they are still influenced by the screening from all layers, leading to complex dynamical interactions between electrons, plasmons and polar phonons. We develop an efficient ab initio model to solve the dynamical electric response of Van der Waals heterostructures, accompanied by a formalism to extract relevant spectroscopic and transport quantities. Notably, we obtain scattering rates for electrons of the heterostructure coupling remotely with electrodynamic excitations. We apply those developments to BN-capped graphene, in which polar phonons from BN couple to plasmons in graphene. We study the nature of the coupled excitations, their dispersion and their coupling to graphene's electrons. Regimes driven by either phonons or plasmons are identified, as well as a truly hybrid regime at long wavelengths. Those are studied as a function of the graphene's Fermi level and the number of BN layers. In contrast with descriptions in terms of surface-optical phonons, we find that the electron-phonon interaction stems from different modes. Moreover, the dynamical screening of the coupling between BN's LO phonons and graphene's electrons crosses over from inefficient to metal-like depending on the relative value of the phonons' frequency and the energetic onset of interband transitions. While the coupling is significant in general, the associated scattering of graphene's carriers is found to be negligible in the context of electronic transport. | [
{
"version": "v1",
"created": "Thu, 21 Mar 2024 11:42:49 GMT"
},
{
"version": "v2",
"created": "Sat, 07 Jun 2025 14:04:20 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Macheda",
"Francesco",
""
],
[
"Mauri",
"Francesco",
""
],
[
"Sohier",
"Thibault",
""
]
] |
2403.14539 | Junhyeong Cho | Junhyeong Cho, Kim Youwang, Hunmin Yang, Tae-Hyun Oh | Robust 3D Shape Reconstruction in Zero-Shot from a Single Image in the Wild | Accepted to CVPR 2025, Project Page: https://ZeroShape-W.github.io | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Recent monocular 3D shape reconstruction methods have shown promising zero-shot results on object-segmented images without any occlusions. However, their effectiveness is significantly compromised in real-world conditions, due to imperfect object segmentation by off-the-shelf models and the prevalence of occlusions. To effectively address these issues, we propose a unified regression model that integrates segmentation and reconstruction, specifically designed for occlusion-aware 3D shape reconstruction. To facilitate its reconstruction in the wild, we also introduce a scalable data synthesis pipeline that simulates a wide range of variations in objects, occluders, and backgrounds. Training on our synthetic data enables the proposed model to achieve state-of-the-art zero-shot results on real-world images, using significantly fewer parameters than competing approaches. | [
{
"version": "v1",
"created": "Thu, 21 Mar 2024 16:40:10 GMT"
},
{
"version": "v2",
"created": "Thu, 28 Nov 2024 13:53:55 GMT"
},
{
"version": "v3",
"created": "Mon, 09 Jun 2025 13:18:25 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Cho",
"Junhyeong",
""
],
[
"Youwang",
"Kim",
""
],
[
"Yang",
"Hunmin",
""
],
[
"Oh",
"Tae-Hyun",
""
]
] |
2403.16706 | Tiejun Tong | Ke Yang, Enxuan Lin, Wangli Xu, Liping Zhu, Tiejun Tong | An alternative measure for quantifying the heterogeneity in meta-analysis | 40 pages, 7 figures and 3 tables | Statistics in Medicine, 44:e70089 (2025) | 10.1002/sim.70089 | null | stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantifying the heterogeneity is an important issue in meta-analysis, and among the existing measures, the $I^2$ statistic is most commonly used. In this paper, we first illustrate with a simple example that the $I^2$ statistic is heavily dependent on the study sample sizes, mainly because it is used to quantify the heterogeneity between the observed effect sizes. To reduce the influence of sample sizes, we introduce an alternative measure that aims to directly measure the heterogeneity between the study populations involved in the meta-analysis. We further propose a new estimator, namely the $I_A^2$ statistic, to estimate the newly defined measure of heterogeneity. For practical implementation, the exact formulas of the $I_A^2$ statistic are also derived under two common scenarios with the effect size as the mean difference (MD) or the standardized mean difference (SMD). Simulations and real data analysis demonstrate that the $I_A^2$ statistic provides an asymptotically unbiased estimator for the absolute heterogeneity between the study populations, and it is also independent of the study sample sizes as expected. To conclude, our newly defined $I_A^2$ statistic can be used as a supplemental measure of heterogeneity to monitor the situations where the study effect sizes are indeed similar with little biological difference. In such scenario, the fixed-effect model can be appropriate; nevertheless, when the sample sizes are sufficiently large, the $I^2$ statistic may still increase to 1 and subsequently suggest the random-effects model for meta-analysis. | [
{
"version": "v1",
"created": "Mon, 25 Mar 2024 12:43:52 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 14:07:55 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Yang",
"Ke",
""
],
[
"Lin",
"Enxuan",
""
],
[
"Xu",
"Wangli",
""
],
[
"Zhu",
"Liping",
""
],
[
"Tong",
"Tiejun",
""
]
] |
2403.20331 | Atsuyuki Miyai | Atsuyuki Miyai, Jingkang Yang, Jingyang Zhang, Yifei Ming, Qing Yu, Go Irie, Yixuan Li, Hai Li, Ziwei Liu, Kiyoharu Aizawa | Unsolvable Problem Detection: Robust Understanding Evaluation for Large Multimodal Models | Accepted by ACL 2025 Main Conference | null | null | null | cs.CV cs.AI cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces a novel task to evaluate the robust understanding capability of Large Multimodal Models (LMMs), termed $\textbf{Unsolvable Problem Detection (UPD)}$. Multiple-choice question answering (MCQA) is widely used to assess the understanding capability of LMMs, but it does not guarantee that LMMs truly comprehend the answer. UPD assesses the LMM's ability to withhold answers when encountering unsolvable problems of MCQA, verifying whether the model truly understands the answer. UPD encompasses three problems: Absent Answer Detection (AAD), Incompatible Answer Set Detection (IASD), and Incompatible Visual Question Detection (IVQD), covering unsolvable cases like answer-lacking or incompatible choices and image-question mismatches. For the evaluation, we introduce the MM-UPD Bench, a benchmark for assessing performance across various ability dimensions. Our experiments reveal that even most LMMs, which demonstrate adequate performance on existing benchmarks, struggle significantly with MM-UPD, underscoring a novel aspect of trustworthiness that current benchmarks have overlooked. A detailed analysis shows that LMMs have different bottlenecks and chain-of-thought and self-reflection improved performance for LMMs with the bottleneck in their LLM capability. We hope our insights will enhance the broader understanding and development of more reliable LMMs. The code is available at https://github.com/AtsuMiyai/UPD. | [
{
"version": "v1",
"created": "Fri, 29 Mar 2024 17:59:53 GMT"
},
{
"version": "v2",
"created": "Wed, 09 Apr 2025 17:13:27 GMT"
},
{
"version": "v3",
"created": "Sun, 27 Apr 2025 19:47:13 GMT"
},
{
"version": "v4",
"created": "Mon, 09 Jun 2025 11:51:00 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Miyai",
"Atsuyuki",
""
],
[
"Yang",
"Jingkang",
""
],
[
"Zhang",
"Jingyang",
""
],
[
"Ming",
"Yifei",
""
],
[
"Yu",
"Qing",
""
],
[
"Irie",
"Go",
""
],
[
"Li",
"Yixuan",
""
],
[
"Li",
"Hai",
""
],
[
"Liu",
"Ziwei",
""
],
[
"Aizawa",
"Kiyoharu",
""
]
] |
2404.00715 | Orlando Luongo | Alessio Belfiglio, Orlando Luongo, Stefano Mancini, and Sebastiano Tomasi | Entanglement entropy in quantum black holes | 15 pages, 7 figures | Class.Quant.Grav. 42 (2025) 3, 035006 | 10.1088/1361-6382/ad9e66 | null | gr-qc hep-th quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We discuss the entanglement entropy for a massive Klein-Gordon field in two Schwarzschild-like quantum black hole spacetimes, also including a nonminimal coupling term with the background scalar curvature. To compute the entanglement entropy, we start from the standard spherical shell discretisation procedure, tracing over the degrees of freedom residing inside an imaginary surface. We estimate the free parameters for such quantum metrics through a simple physical argument based on Heisenberg uncertainty principle, along with alternative proposals as asymptotic safety, trace anomaly, and graviton corpuscular scaling. Our findings reveal a significant decrease in entropy compared to the area law near the origin for the quantum metrics. In both scenarios, the entanglement entropy converges to the expected area law sufficiently far from the origin. We then compare these results to the entropy scaling in regular Hayward and corrected-Hayward spacetimes to highlight the main differences with such regular approaches. | [
{
"version": "v1",
"created": "Sun, 31 Mar 2024 15:19:03 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Belfiglio",
"Alessio",
""
],
[
"Luongo",
"Orlando",
""
],
[
"Mancini",
"Stefano",
""
],
[
"Tomasi",
"Sebastiano",
""
]
] |
2404.01949 | Qizhi Qiu | Qizhi Qiu, Xiaomin Liu, Yihao Zhang, Lilin Yi, Weisheng Hu, Qunbi Zhuge | Heuristic Optimization of Amplifier Reconfiguration Process for Autonomous Driving Optical Networks | null | ECOC 2024; 50th European Conference on Optical Communication, Frankfurt, Germany, 22-26 September 2024, pp. 152-155 | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a heuristic-based optimization scheme for reliable optical amplifier reconfiguration process in ADON. In the experiment on a commercial testbed, the scheme prevents a 1.0-dB Q-factor degradation and outperforms 98.5% random solutions. | [
{
"version": "v1",
"created": "Tue, 02 Apr 2024 13:44:27 GMT"
},
{
"version": "v2",
"created": "Fri, 19 Jul 2024 01:24:16 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Qiu",
"Qizhi",
""
],
[
"Liu",
"Xiaomin",
""
],
[
"Zhang",
"Yihao",
""
],
[
"Yi",
"Lilin",
""
],
[
"Hu",
"Weisheng",
""
],
[
"Zhuge",
"Qunbi",
""
]
] |
2404.02791 | Thomas Ouldridge | Benjamin Qureshi, Jenny M. Poulton and Thomas E. Ouldridge | Thermodynamic limits on general far-from-equilibrium molecular templating networks | null | null | null | null | physics.bio-ph cond-mat.stat-mech cs.IT math.IT q-bio.MN | http://creativecommons.org/licenses/by/4.0/ | Cells produce RNA and proteins via molecular templating networks. We show that information transmission in such networks is bounded by functions of a simple thermodynamic property of the network, regardless of complexity. Surprisingly, putative systems operating at this bound do not have a high flux around the network. Instead, they have low entropy production, with each product in a ``pseudo-equilibrium'' determined by a single pathway. These pseudo-equilibrium limits constrain information transmission for the overall network, even if individual templates are arbitrarily specific. | [
{
"version": "v1",
"created": "Wed, 03 Apr 2024 14:58:30 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 12:31:13 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Qureshi",
"Benjamin",
""
],
[
"Poulton",
"Jenny M.",
""
],
[
"Ouldridge",
"Thomas E.",
""
]
] |
2404.02800 | Bernardo Leite | Bernardo Leite and Henrique Lopes Cardoso | On Few-Shot Prompting for Controllable Question-Answer Generation in Narrative Comprehension | Preprint - Accepted for publication at CSEDU 2024 | CSEDU 2024 Proceedings, SCITEPRESS, 2024 | 10.5220/0012623800003693 | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Question Generation aims to automatically generate questions based on a given input provided as context. A controllable question generation scheme focuses on generating questions with specific attributes, allowing better control. In this study, we propose a few-shot prompting strategy for controlling the generation of question-answer pairs from children's narrative texts. We aim to control two attributes: the question's explicitness and underlying narrative elements. With empirical evaluation, we show the effectiveness of controlling the generation process by employing few-shot prompting side by side with a reference model. Our experiments highlight instances where the few-shot strategy surpasses the reference model, particularly in scenarios such as semantic closeness evaluation and the diversity and coherency of question-answer pairs. However, these improvements are not always statistically significant. The code is publicly available at github.com/bernardoleite/few-shot-prompting-qg-control. | [
{
"version": "v1",
"created": "Wed, 03 Apr 2024 15:17:21 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Leite",
"Bernardo",
""
],
[
"Cardoso",
"Henrique Lopes",
""
]
] |
2404.03108 | Nikhita Joshi | Nikhita Joshi and Daniel Vogel | Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help | Accepted to the 7th ACM Conference on Conversational User Interfaces (CUI '25) | null | 10.1145/3719160.3736608 | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The feeling of something belonging to someone is called "psychological ownership." A common assumption is that writing with generative AI lowers psychological ownership, but the extent to which this occurs and the role of prompt length are unclear. We report on two experiments to examine the relationship between psychological ownership and prompt length. Participants wrote short stories either completely by themselves or wrote prompts of varying lengths. Results show that when participants wrote longer prompts, they had higher levels of psychological ownership. Their comments suggest they thought more about their prompts, often adding more details about the plot. However, benefits plateaued when prompt length was 75-100% of the target story length. To encourage users to write longer prompts, we propose augmenting the prompt submission button so it must be held down a long time if the prompt is short. Results show that this technique is effective at increasing prompt length. | [
{
"version": "v1",
"created": "Wed, 03 Apr 2024 23:19:45 GMT"
},
{
"version": "v2",
"created": "Thu, 22 Aug 2024 18:36:17 GMT"
},
{
"version": "v3",
"created": "Thu, 20 Feb 2025 16:46:10 GMT"
},
{
"version": "v4",
"created": "Sun, 08 Jun 2025 19:39:53 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Joshi",
"Nikhita",
""
],
[
"Vogel",
"Daniel",
""
]
] |
2404.03283 | Anna Franziska Michael | Anna Michael, Yuri Santos Rego, Petra Schwer, Olga Varghese | Involutions in Coxeter groups | 31 pages, 12 figures | Algebras and Representation Theory, volume 28 (2025), 647-667 | 10.1007/s10468-025-10332-x | null | math.GR math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We combinatorially characterize the number $\mathrm{cc}_2$ of conjugacy classes of involutions in any Coxeter group in terms of higher rank odd graphs. This notion naturally generalizes the concept of odd graphs, used previously to count the number of conjugacy classes of reflections. We provide uniform bounds and discuss some extremal cases, where the number $\mathrm{cc}_2$ is smallest or largest possible. Moreover, we provide formulae for $\mathrm{cc}_2$ in free and direct products as well as for some finite and affine types, besides computing $\mathrm{cc}_2$ for all triangle groups, and all affine irreducible Coxeter groups of rank up to eleven. | [
{
"version": "v1",
"created": "Thu, 04 Apr 2024 08:12:38 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Michael",
"Anna",
""
],
[
"Rego",
"Yuri Santos",
""
],
[
"Schwer",
"Petra",
""
],
[
"Varghese",
"Olga",
""
]
] |
2404.03900 | Jerry Yao-Chieh Hu | Jerry Yao-Chieh Hu, Bo-Yu Chen, Dennis Wu, Feng Ruan, Han Liu | Nonparametric Modern Hopfield Models | Accepted at ICML 2025. Code available at https://github.com/MAGICS-LAB/NonparametricHopfield. v2 matches with camera-ready version | null | null | null | stat.ML cs.AI cs.LG cs.NE | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present a nonparametric interpretation for deep learning compatible modern Hopfield models and utilize this new perspective to debut efficient variants. Our key contribution stems from interpreting the memory storage and retrieval processes in modern Hopfield models as a nonparametric regression problem subject to a set of query-memory pairs. Interestingly, our framework not only recovers the known results from the original dense modern Hopfield model but also fills the void in the literature regarding efficient modern Hopfield models, by introducing \textit{sparse-structured} modern Hopfield models with sub-quadratic complexity. We establish that this sparse model inherits the appealing theoretical properties of its dense analogue -- connection with transformer attention, fixed point convergence and exponential memory capacity. Additionally, we showcase the versatility of our framework by constructing a family of modern Hopfield models as extensions, including linear, random masked, top-$K$ and positive random feature modern Hopfield models. Empirically, we validate our framework in both synthetic and realistic settings for memory retrieval and learning tasks. | [
{
"version": "v1",
"created": "Fri, 05 Apr 2024 05:46:20 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 03:41:40 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Hu",
"Jerry Yao-Chieh",
""
],
[
"Chen",
"Bo-Yu",
""
],
[
"Wu",
"Dennis",
""
],
[
"Ruan",
"Feng",
""
],
[
"Liu",
"Han",
""
]
] |
2404.04074 | Soham Mukherjee | Soham Mukherjee, Manfred Claassen, Paul-Christian B\"urkner | DGP-LVM: Derivative Gaussian process latent variable models | 36 pages, 33 figures | null | 10.1007/s11222-025-10644-4 | null | stat.ME | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We develop a framework for derivative Gaussian process latent variable models (DGP-LVMs) that can handle multi-dimensional output data using modified derivative covariance functions. The modifications account for complexities in the underlying data generating process such as scaled derivatives, varying information across multiple output dimensions as well as interactions between outputs. Further, our framework provides uncertainty estimates for each latent variable samples using Bayesian inference. Through extensive simulations, we demonstrate that latent variable estimation accuracy can be drastically increased by including derivative information due to our proposed covariance function modifications. The developments are motivated by a concrete biological research problem involving the estimation of the unobserved cellular ordering from single-cell RNA (scRNA) sequencing data for gene expression and its corresponding derivative information known as RNA velocity. Since the RNA velocity is only an estimate of the exact derivative information, the derivative covariance functions need to account for potential scale differences. In a real-world case study, we illustrate the application of DGP-LVMs to such scRNA sequencing data. While motivated by this biological problem, our framework is generally applicable to all kinds of latent variable estimation problems involving derivative information irrespective of the field of study. | [
{
"version": "v1",
"created": "Fri, 05 Apr 2024 13:03:13 GMT"
},
{
"version": "v2",
"created": "Tue, 09 Jul 2024 16:56:07 GMT"
},
{
"version": "v3",
"created": "Fri, 31 Jan 2025 16:01:37 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Mukherjee",
"Soham",
""
],
[
"Claassen",
"Manfred",
""
],
[
"Bürkner",
"Paul-Christian",
""
]
] |
2404.04166 | Evan O'Dorney | Evan M. O'Dorney | Even-carry polynomials and cohomology of line bundles on the incidence correspondence in positive characteristic | 16 pages, including 4 tables | Experimental Mathematics, 1-14, 2025 | 10.1080/10586458.2025.2481257 | null | math.AG math.AC math.NT math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the cohomology groups of line bundles $\mathcal{L}$ on the \emph{incidence correspondence}, that is, a general hypersurface $X \subset \mathbb{P}^{n-1} \times \mathbb{P}^{n-1}$ of degrees $(1,1)$. Whereas the characteristic $0$ situation is completely understood, the cohomology in characteristic $p$ depends in a mysterious way on the base-$p$ digits of the degrees $(d, e)$ of $\mathcal{L}$. Gao and Raicu (following Linyuan Liu) prove a recursive description of the cohomology for $n = 3$, which relates to Nim polynomials when $p = 2$. In this paper, we devise a suitable generalization of Nim polynomials, which we call \emph{even-carry polynomials,} by which we can solve the recurrence of Liu--Gao--Raicu to yield an explicit formula for the cohomology for $n = 3$ and general $p$. We also make some conjectures on the general form of the cohomology for general $n$ and $p$. | [
{
"version": "v1",
"created": "Fri, 05 Apr 2024 15:19:54 GMT"
},
{
"version": "v2",
"created": "Wed, 21 Aug 2024 18:26:33 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"O'Dorney",
"Evan M.",
""
]
] |
2404.04218 | Filip Koprivec | Filip Koprivec and Matija Pretnar | Simplifying explicit subtyping coercions in a polymorphic calculus with effects | 39 pages, 11 figures | null | null | null | cs.PL | http://creativecommons.org/licenses/by/4.0/ | Algebraic effect handlers are becoming an increasingly popular way of structuring effectful computations, and their performance is often a concern. One of the proposed approaches towards efficient compilation is tracking effect information through explicit subtyping coercions. However, in the presence of polymorphism, these coercions are compiled into additional arguments of compiled functions, incurring significant overhead. In this paper, we present a polymorphic effectful calculus, identify simplification phases needed to reduce the number of unnecessary constraints, and prove that they preserve semantics. In addition, we implement the simplification algorithm in the Eff language and evaluate its performance on a number of benchmarks. Though we do not prove the optimality of the presented simplifications, the results show that the algorithm eliminates all coercions, resulting in code as efficient as manually monomorphised one. | [
{
"version": "v1",
"created": "Fri, 05 Apr 2024 17:04:17 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Jun 2025 21:43:27 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Koprivec",
"Filip",
""
],
[
"Pretnar",
"Matija",
""
]
] |
2404.04411 | Jan Balewski | Jan Balewski, Milan Kornja\v{c}a, Katherine Klymko, Siva Darbha, Mark R. Hirsbrunner, Pedro L. S. Lopes, Fangli Liu, Daan Camps | Engineering quantum states with neutral atoms | 7 pages, 6 figures, IEEE QCE24 - QTEM Best Paper Award - 1st Place | null | 10.1109/QCE60285.2024.00144 | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Aquila, an analog quantum simulation platform developed by QuEra Computing, supports control of the position and coherent evolution of up to 256 neutral atoms. This study details novel experimental protocols designed for analog quantum simulators that generate Bell state entanglement far away from the blockade regime, construct a $Z_2$ state with a defect induced by an ancilla, and optimize the driving fields schedule to prepare excited states with enhanced fidelity. We additionally evaluate the effectiveness of readout error mitigation techniques in improving the fidelity of measurement results. All experiments were executed on Aquila from QuEra and facilitated by the AWS Braket interface. Our experimental results closely align with theoretical predictions and numerical simulations. The insights gained from this study showcase Aquila's capabilities in handling complex quantum simulations and computations, and also pave the way for new avenues of research in quantum information processing and physics that employ programmable analog hardware platforms. | [
{
"version": "v1",
"created": "Fri, 05 Apr 2024 21:20:41 GMT"
},
{
"version": "v2",
"created": "Fri, 04 Oct 2024 22:38:13 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Balewski",
"Jan",
""
],
[
"Kornjača",
"Milan",
""
],
[
"Klymko",
"Katherine",
""
],
[
"Darbha",
"Siva",
""
],
[
"Hirsbrunner",
"Mark R.",
""
],
[
"Lopes",
"Pedro L. S.",
""
],
[
"Liu",
"Fangli",
""
],
[
"Camps",
"Daan",
""
]
] |
2404.04656 | Seungjae Jung | Seungjae Jung, Gunsoo Han, Daniel Wontae Nam and Kyoung-Woon On | Binary Classifier Optimization for Large Language Model Alignment | ACL 2025 main | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | In real-world services such as ChatGPT, aligning models based on user feedback is crucial for improving model performance. However, due to the simplicity and convenience of providing feedback, users typically offer only basic binary signals, such as 'thumbs-up' or 'thumbs-down'. Most existing alignment research, on the other hand, relies on preference-based approaches that require both positive and negative responses as a pair. We propose Binary Classifier Optimization (BCO), a technique that effectively aligns LLMs using only binary feedback. BCO trains a binary classifier, where the logit serves as an implicit reward, effectively minimizing the Direct Preference Optimization (DPO) loss. We demonstrate that the binary cross-entropy loss employed in classifier training acts as an upper bound for the DPO loss. Additionally, a novel reward shift technique further minimizes the gap between the losses. We validate our methodology in two settings: first, on a paired preference dataset, where our method performs on par with DPO; and second, on a Likert-5 scale annotation dataset which stems from real users' queries. Our model consistently demonstrates effective and robust alignment across four base LLMs and three different datasets, showcasing the strength of our approach to learning from binary signals. | [
{
"version": "v1",
"created": "Sat, 06 Apr 2024 15:20:59 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 07:10:33 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Jung",
"Seungjae",
""
],
[
"Han",
"Gunsoo",
""
],
[
"Nam",
"Daniel Wontae",
""
],
[
"On",
"Kyoung-Woon",
""
]
] |
2404.05678 | Yuheng Lai | Yuheng Lai, Leying Guan | FairICP: Encouraging Equalized Odds via Inverse Conditional Permutation | null | null | null | null | stat.ML cs.CY cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | $\textit{Equalized odds}$, an important notion of algorithmic fairness, aims to ensure that sensitive variables, such as race and gender, do not unfairly influence the algorithm's prediction when conditioning on the true outcome. Despite rapid advancements, current research primarily focuses on equalized odds violations caused by a single sensitive attribute, leaving the challenge of simultaneously accounting for multiple attributes under-addressed. We bridge this gap by introducing an in-processing fairness-aware learning approach, FairICP, which integrates adversarial learning with a novel inverse conditional permutation scheme. FairICP offers a flexible and efficient scheme to promote equalized odds under fairness conditions described by complex and multi-dimensional sensitive attributes. The efficacy and adaptability of our method are demonstrated through both simulation studies and empirical analyses of real-world datasets. | [
{
"version": "v1",
"created": "Mon, 08 Apr 2024 16:57:44 GMT"
},
{
"version": "v2",
"created": "Tue, 09 Apr 2024 16:17:14 GMT"
},
{
"version": "v3",
"created": "Fri, 01 Nov 2024 16:03:23 GMT"
},
{
"version": "v4",
"created": "Fri, 06 Jun 2025 18:36:08 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Lai",
"Yuheng",
""
],
[
"Guan",
"Leying",
""
]
] |
2404.08210 | Frederik Geth | Ching Hong Tam, Frederik Geth, Nadarajah Mithulananthan | The Inverse Carson's Equations Problem: Definition, Implementation and Numerical Experiments | 12 pages | null | null | null | math.OC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In recent years, with the increase in renewable energy and storage penetration, power flow studies in low-voltage networks have become of interest in both industry and academia. Many studies use impedance represented by sequence components due to the lack of datasets with fully parameterized impedance matrices. This assumes that the network impedance is balanced, which is typically not the case in the low-voltage network and therefore risks the accuracy of the study. This paper proposes a methodology for the recovery of more detailed impedance data from sequence components as an inverse problem, i.e. the inverse Carson's equations problem, for both overhead lines and cables. We consider discrete properties like material and configuration of conductors common in the distribution network and investigate what data can be reliably recovered from only sequence components using nonlinear optimisation models. Presented results include uniqueness of recovered variables and the likelihood of mismatch. | [
{
"version": "v1",
"created": "Fri, 12 Apr 2024 02:48:33 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 03:42:13 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Tam",
"Ching Hong",
""
],
[
"Geth",
"Frederik",
""
],
[
"Mithulananthan",
"Nadarajah",
""
]
] |
2404.08578 | Pat Lank | Alexander Clark, Pat Lank, Kabeer Manali-Rahul, Chris J. Parker | Classification and nonexistence for $t$-structures on derived categories of schemes | Current: Minor typos, added details. Previous: Significant improvement to prior results, expanded exposition | null | null | null | math.AG math.CT | http://creativecommons.org/licenses/by/4.0/ | This work establishes new results on the classification of $t$-structures for many subcategories of the derived category of quasi-coherent sheaves on a Noetherian scheme. Our work makes progress in two different directions. On one hand, we provide an improvement of a result of Takahashi on $t$-structures, generalizing it to the case of the bounded derived category of coherent sheaves on a quasi-compact CM-excellent scheme of finite Krull dimension. On the other hand, via independent techniques, we prove a variation of a recent result of Neeman which resolved a conjecture of Antieau, Gepner, and Heller. | [
{
"version": "v1",
"created": "Fri, 12 Apr 2024 16:23:21 GMT"
},
{
"version": "v2",
"created": "Tue, 13 Aug 2024 12:03:36 GMT"
},
{
"version": "v3",
"created": "Wed, 08 Jan 2025 15:44:15 GMT"
},
{
"version": "v4",
"created": "Sat, 07 Jun 2025 07:39:40 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Clark",
"Alexander",
""
],
[
"Lank",
"Pat",
""
],
[
"Manali-Rahul",
"Kabeer",
""
],
[
"Parker",
"Chris J.",
""
]
] |
2404.08606 | Mark Lawson | Mark V. Lawson | The structure of \'etale Boolean right restriction monoids | Paper revised; typos corrected; added more on Cantor algebras | null | null | null | math.CT | http://creativecommons.org/licenses/by/4.0/ | In this paper, we describe \'etale Boolean right restriction monoids in terms of Boolean inverse monoids. | [
{
"version": "v1",
"created": "Fri, 12 Apr 2024 17:09:52 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 09:43:24 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Lawson",
"Mark V.",
""
]
] |
2404.08634 | Sunny Sanyal | Sunny Sanyal, Ravid Shwartz-Ziv, Alexandros G. Dimakis, Sujay Sanghavi | When Attention Collapses: How Degenerate Layers in LLMs Enable Smaller, Stronger Models | 29 pages, 22 figures, 11 tables | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) rely on the transformer architecture and its self-attention mechanism to deliver strong performance across tasks. However, we uncover a structural inefficiency in standard pre-trained decoder-style LLMs: in many of the deeper layers, attention matrices frequently collapse to near rank-one, single-column patterns. We refer to these underutilized components as lazy layers, which are redundant and computationally inefficient. To address this, we propose Inheritune, a simple and effective training recipe for building smaller, more efficient, and high performing language models. Inheritune initializes a compact model by inheriting the useful early layers from a larger pre-trained model, then progressively retrains and expands it. Our experiments across multiple models and datasets show that Inheritune trained models, despite having significantly fewer layers, can match or even outperform their larger counterparts. This approach yields compact, performant models and offers a practical path for efficient language model compression. Code is available at https://github.com/sanyalsunny111/LLM-Inheritune | [
{
"version": "v1",
"created": "Fri, 12 Apr 2024 17:53:34 GMT"
},
{
"version": "v2",
"created": "Fri, 04 Oct 2024 05:14:48 GMT"
},
{
"version": "v3",
"created": "Sun, 08 Jun 2025 09:19:32 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Sanyal",
"Sunny",
""
],
[
"Shwartz-Ziv",
"Ravid",
""
],
[
"Dimakis",
"Alexandros G.",
""
],
[
"Sanghavi",
"Sujay",
""
]
] |
2404.08793 | Yingchaojie Feng | Yingchaojie Feng, Zhizhang Chen, Zhining Kang, Sijia Wang, Haoyu Tian, Wei Zhang, Minfeng Zhu, Wei Chen | JailbreakLens: Visual Analysis of Jailbreak Attacks Against Large Language Models | null | null | 10.1109/TVCG.2025.3575694 | null | cs.CR cs.CL cs.HC | http://creativecommons.org/licenses/by/4.0/ | The proliferation of large language models (LLMs) has underscored concerns regarding their security vulnerabilities, notably against jailbreak attacks, where adversaries design jailbreak prompts to circumvent safety mechanisms for potential misuse. Addressing these concerns necessitates a comprehensive analysis of jailbreak prompts to evaluate LLMs' defensive capabilities and identify potential weaknesses. However, the complexity of evaluating jailbreak performance and understanding prompt characteristics makes this analysis laborious. We collaborate with domain experts to characterize problems and propose an LLM-assisted framework to streamline the analysis process. It provides automatic jailbreak assessment to facilitate performance evaluation and support analysis of components and keywords in prompts. Based on the framework, we design JailbreakLens, a visual analysis system that enables users to explore the jailbreak performance against the target model, conduct multi-level analysis of prompt characteristics, and refine prompt instances to verify findings. Through a case study, technical evaluations, and expert interviews, we demonstrate our system's effectiveness in helping users evaluate model security and identify model weaknesses. | [
{
"version": "v1",
"created": "Fri, 12 Apr 2024 19:54:42 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Jun 2025 20:39:01 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Feng",
"Yingchaojie",
""
],
[
"Chen",
"Zhizhang",
""
],
[
"Kang",
"Zhining",
""
],
[
"Wang",
"Sijia",
""
],
[
"Tian",
"Haoyu",
""
],
[
"Zhang",
"Wei",
""
],
[
"Zhu",
"Minfeng",
""
],
[
"Chen",
"Wei",
""
]
] |
2404.09480 | Kyubyung Chae | Kyubyung Chae, Jaepill Choi, Yohan Jo, Taesup Kim | Mitigating Hallucination in Abstractive Summarization with Domain-Conditional Mutual Information | Accepted by Findings of NAACL 2024 | null | 10.18653/v1/2024.findings-naacl.117 | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A primary challenge in abstractive summarization is hallucination -- the phenomenon where a model generates plausible text that is absent in the source text. We hypothesize that the domain (or topic) of the source text triggers the model to generate text that is highly probable in the domain, neglecting the details of the source text. To alleviate this model bias, we introduce a decoding strategy based on domain-conditional pointwise mutual information. This strategy adjusts the generation probability of each token by comparing it with the token's marginal probability within the domain of the source text. According to evaluation on the XSUM dataset, our method demonstrates improvement in terms of faithfulness and source relevance. The code is publicly available at \url{https://github.com/qqplot/dcpmi}. | [
{
"version": "v1",
"created": "Mon, 15 Apr 2024 06:06:43 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Chae",
"Kyubyung",
""
],
[
"Choi",
"Jaepill",
""
],
[
"Jo",
"Yohan",
""
],
[
"Kim",
"Taesup",
""
]
] |
2404.10426 | Cristian Urbina | Sara Giuliani, Shunsuke Inenaga, Zsuzsanna Lipt\'ak, Giuseppe Romana, Marinella Sciortino and Cristian Urbina | Bit catastrophes for the Burrows-Wheeler Transform | This preprint is an extended version of our conference article with the same title, published in the proceedings of DLT 2023. Full version of this preprint can be found at Theory of Computing Systems: https://link.springer.com/article/10.1007/s00224-024-10212-9 | Theory of Computing Systems 69, 19 (2025) | 10.1007/s00224-024-10212-9 | null | cs.DS cs.DM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A bit catastrophe, loosely defined, is when a change in just one character of a string causes a significant change in the size of the compressed string. We study this phenomenon for the Burrows-Wheeler Transform (BWT), a string transform at the heart of several of the most popular compressors and aligners today. The parameter determining the size of the compressed data is the number of equal-letter runs of the BWT, commonly denoted $r$.
We exhibit infinite families of strings in which insertion, deletion, resp. substitution of one character increases $r$ from constant to $\Theta(\log n)$, where $n$ is the length of the string. These strings can be interpreted both as examples for an increase by a multiplicative or an additive $\Theta(\log n)$-factor. As regards multiplicative factor, they attain the upper bound given by Akagi, Funakoshi, and Inenaga [Inf & Comput. 2023] of $O(\log n \log r)$, since here $r=O(1)$.
We then give examples of strings in which insertion, deletion, resp. substitution of a character increases $r$ by a $\Theta(\sqrt{n})$ additive factor. These strings significantly improve the best known lower bound for an additive factor of $\Omega(\log n)$ [Giuliani et al., SOFSEM 2021]. | [
{
"version": "v1",
"created": "Tue, 16 Apr 2024 09:44:14 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 01:11:08 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Giuliani",
"Sara",
""
],
[
"Inenaga",
"Shunsuke",
""
],
[
"Lipták",
"Zsuzsanna",
""
],
[
"Romana",
"Giuseppe",
""
],
[
"Sciortino",
"Marinella",
""
],
[
"Urbina",
"Cristian",
""
]
] |
2404.11871 | Hanjing Zhou | Hanjing Zhou, Mingze Yin, Danny Chen, Jian Wu, JinTai Chen | Group-On: Boosting One-Shot Segmentation with Supportive Query | null | ICME 2025 | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One-shot semantic segmentation aims to segment query images given only ONE annotated support image of the same class. This task is challenging because target objects in the support and query images can be largely different in appearance and pose (i.e., intra-class variation). Prior works suggested that incorporating more annotated support images in few-shot settings boosts performances but increases costs due to additional manual labeling. In this paper, we propose a novel and effective approach for ONE-shot semantic segmentation, called Group-On, which packs multiple query images in batches for the benefit of mutual knowledge support within the same category. Specifically, after coarse segmentation masks of the batch of queries are predicted, query-mask pairs act as pseudo support data to enhance mask predictions mutually. To effectively steer such process, we construct an innovative MoME module, where a flexible number of mask experts are guided by a scene-driven router and work together to make comprehensive decisions, fully promoting mutual benefits of queries. Comprehensive experiments on three standard benchmarks show that, in the ONE-shot setting, Group-On significantly outperforms previous works by considerable margins. With only one annotated support image, Group-On can be even competitive with the counterparts using 5 annotated images. | [
{
"version": "v1",
"created": "Thu, 18 Apr 2024 03:10:04 GMT"
},
{
"version": "v2",
"created": "Sun, 08 Jun 2025 12:11:23 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Zhou",
"Hanjing",
""
],
[
"Yin",
"Mingze",
""
],
[
"Chen",
"Danny",
""
],
[
"Wu",
"Jian",
""
],
[
"Chen",
"JinTai",
""
]
] |
2404.12206 | Roman Kuzian | R.O. Kuzian, E. E. Krasovskii | Attosecond dynamics of electron scattering by an absorbing layer | 8 pages, 5 figures | null | 10.1016/j.elspec.2025.147535 | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Attosecond dynamics of electron reflection from a thin film is studied based on a one-dimensional jellium model. Following the Eisenbud-Wigner-Smith concept, the reflection time delay $\Delta\tau_{\rm R}$ is calculated as the energy derivative of the phase of the complex reflection amplitude $r$. For a purely elastic scattering by a jellium slab of a finite thickness $d$ the transmission probability $T$ oscillates with the momentum $K$ in the solid with a period $\pi/d$, and $\Delta\tau_{\rm R}$ closely follows these oscillations. The reflection delay averaged over an energy interval grows with $d$, but in the limit of $d\to\infty$ the amplitude $r$ becomes real, so $\Delta\tau_{\rm R}$ vanishes. This picture changes substantially with the inclusion of an absorbing potential $-iV_{\rm i}$: As expected, for a sufficiently thick slab the reflection amplitude now tends to its asymptotic value for a semi-infinite crystal. Interestingly, for $V_{\rm i} \ne 0$, around the $T(E)$ maxima, the $\Delta\tau_{\rm R}(E)$ curve strongly deviates from $T(E)$, showing a narrow dip just at the $\Delta\tau_{\rm R}(E)$ maximum for $V_{\rm i}=0$. An analytical theory of this counterintuitive behavior is developed. | [
{
"version": "v1",
"created": "Thu, 18 Apr 2024 14:12:05 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Kuzian",
"R. O.",
""
],
[
"Krasovskii",
"E. E.",
""
]
] |
2404.12606 | Yu-Hsuan Ho | Yu-Hsuan Ho, Longxiang Li, Ali Mostafavi | ELEV-VISION-SAM: Integrated Vision Language and Foundation Model for Automated Estimation of Building Lowest Floor Elevation | null | Comput. Aided Civ. Infrastruct. Eng. 40.1 (2025) 75-90 | 10.1111/mice.13310 | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Street view imagery, aided by advancements in image quality and accessibility, has emerged as a valuable resource for urban analytics research. Recent studies have explored its potential for estimating lowest floor elevation (LFE), offering a scalable alternative to traditional on-site measurements, crucial for assessing properties' flood risk and damage extent. While existing methods rely on object detection, the introduction of image segmentation has broadened street view images' utility for LFE estimation, although challenges still remain in segmentation quality and capability to distinguish front doors from other doors. To address these challenges in LFE estimation, this study integrates the Segment Anything model, a segmentation foundation model, with vision language models to conduct text-prompt image segmentation on street view images for LFE estimation. By evaluating various vision language models, integration methods, and text prompts, we identify the most suitable model for street view image analytics and LFE estimation tasks, thereby improving the availability of the current LFE estimation model based on image segmentation from 33% to 56% of properties. Remarkably, our proposed method significantly enhances the availability of LFE estimation to almost all properties in which the front door is visible in the street view image. Also the findings present the first baseline and comparison of various vision models of street view image-based LFE estimation. The model and findings not only contribute to advancing street view image segmentation for urban analytics but also provide a novel approach for image segmentation tasks for other civil engineering and infrastructure analytics tasks. | [
{
"version": "v1",
"created": "Fri, 19 Apr 2024 03:16:08 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Ho",
"Yu-Hsuan",
""
],
[
"Li",
"Longxiang",
""
],
[
"Mostafavi",
"Ali",
""
]
] |
2404.12940 | Grigory Bartosh | Grigory Bartosh, Dmitry Vetrov, Christian A. Naesseth | Neural Flow Diffusion Models: Learnable Forward Process for Improved Diffusion Modelling | null | null | null | null | stat.ML cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Conventional diffusion models typically relies on a fixed forward process, which implicitly defines complex marginal distributions over latent variables. This can often complicate the reverse process' task in learning generative trajectories, and results in costly inference for diffusion models. To address these limitations, we introduce Neural Flow Diffusion Models (NFDM), a novel framework that enhances diffusion models by supporting a broader range of forward processes beyond the standard Gaussian. We also propose a novel parameterization technique for learning the forward process. Our framework provides an end-to-end, simulation-free optimization objective, effectively minimizing a variational upper bound on the negative log-likelihood. Experimental results demonstrate NFDM's strong performance, evidenced by state-of-the-art likelihood estimation. Furthermore, we investigate NFDM's capacity for learning generative dynamics with specific characteristics, such as deterministic straight lines trajectories, and demonstrate how the framework may be adopted for learning bridges between two distributions. The results underscores NFDM's versatility and its potential for a wide range of applications. | [
{
"version": "v1",
"created": "Fri, 19 Apr 2024 15:10:54 GMT"
},
{
"version": "v2",
"created": "Sat, 01 Jun 2024 10:25:54 GMT"
},
{
"version": "v3",
"created": "Fri, 06 Jun 2025 12:38:36 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Bartosh",
"Grigory",
""
],
[
"Vetrov",
"Dmitry",
""
],
[
"Naesseth",
"Christian A.",
""
]
] |
2404.15996 | Seyed Majid Zahedi | Pouya Kananian, Arnesh Sujanani, Seyed Majid Zahedi | Asymptotically Fair and Truthful Allocation of Public Goods | 26 pages, typos corrected, section 4.3 added, new author added | null | null | null | cs.GT | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We study the fair and truthful allocation of m divisible public items among n agents, each with distinct preferences for the items. To aggregate agents' preferences fairly, we focus on finding a core solution. For divisible items, a core solution always exists and can be calculated by maximizing the Nash welfare objective. However, such a solution is easily manipulated; agents might have incentives to misreport their preferences. To mitigate this, the current state-of-the-art finds an approximate core solution with high probability while ensuring approximate truthfulness. However, this approach has two main limitations. First, due to several approximations, the approximation error in the core could grow with n, resulting in a non-asymptotic core solution. This limitation is particularly significant as public-good allocation mechanisms are frequently applied in scenarios involving a large number of agents, such as the allocation of public tax funds for municipal projects. Second, implementing the current approach for practical applications proves to be a highly nontrivial task. To address these limitations, we introduce PPGA, a (differentially) Private Public-Good Allocation algorithm, and show that it attains asymptotic truthfulness and finds an asymptotic core solution with high probability. Additionally, to demonstrate the practical applicability of our algorithm, we implement PPGA and empirically study its properties using municipal participatory budgeting data. | [
{
"version": "v1",
"created": "Wed, 24 Apr 2024 17:15:19 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Jun 2025 19:44:54 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Kananian",
"Pouya",
""
],
[
"Sujanani",
"Arnesh",
""
],
[
"Zahedi",
"Seyed Majid",
""
]
] |
2404.16575 | Jun-Zhang Wang | Jun-Zhang Wang, Zi-Yang Lin, Yan-Ke Chen, Lu Meng and Shi-Lin Zhu | Uncovering the mystery of $X(3872)$ with the coupled-channel dynamics | 8+8 pages, 3+5 figures, added new results and discussion: the successful prediction for new charmoniumlike chi_c1(4010) and its implication for uncovering the pole origin of X(3872), accepted for publication as a Letter in Phys. Rev. D | null | null | null | hep-ph hep-ex hep-lat | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The $X(3872)$, as the first and the most crucial member in the exotic charmoniumlike $XYZ$ family, has been studied for a long time. However, its dynamical origin, whether stemming from a $D\bar{D}^*$ hadronic molecule or the first excited $P$-wave charmonium $\chi_{c1}(2P)$, remains controversial. In this Letter, we demonstrate that the $X(3872)$ definitely does not result from the mass shift of the higher bare $\chi_{c1}(2P)$ resonance pole in the coupled-channel dynamics involving a short-distance $c\bar{c}$ core and the long-distance $D\bar{D}^*$ channels. Instead, it originates from either the $D\bar{D}^*$ molecular pole or the shadow pole associated with the anti-resonance of the $P$-wave charmonium, depending on the weak or strong coupling mode, respectively. To differentiate these origins and fully exploit the nature of $X(3872)$, we conduct a comprehensive analysis in a couple-channel dynamics framework, including the isospin violation, the three-body $D\bar{D}\pi$ effect, the dynamical width of $D^*$, and non-open-charm decays of the bare $\chi_{c1}(2P)$. Our findings highlight the pivotal role of the coupled-channel dynamics in explaining the disparity between the pole widths of $X(3872)$ and $T_{cc}^+$, while also predicting a new resonance with $J^{PC}=1^{++}$ around 4.0 GeV. By matching the newly observed $\chi_{c1}(4010)$ by the LHCb Collaboration to our predicted resonance, we conclude that the $X(3872)$ most likely originates from the $D\bar{D}^*$ pole with a confidence level exceeding $99.7\%$. | [
{
"version": "v1",
"created": "Thu, 25 Apr 2024 12:42:48 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 07:31:20 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Wang",
"Jun-Zhang",
""
],
[
"Lin",
"Zi-Yang",
""
],
[
"Chen",
"Yan-Ke",
""
],
[
"Meng",
"Lu",
""
],
[
"Zhu",
"Shi-Lin",
""
]
] |
2404.18071 | Nishant Luitel | Nishant Luitel, Nirajan Bekoju, Anand Kumar Sah, Subarna Shakya | Can Perplexity Predict Fine-tuning Performance? An Investigation of Tokenization Effects on Sequential Language Models for Nepali | 11 pages | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | The impact of subword tokenization on language model performance is well-documented for perplexity, with finer granularity consistently reducing this intrinsic metric. However, research on how different tokenization schemes affect a model's understanding capabilities remains limited, particularly for non-Latin script languages. Addressing this gap, we conducted a comprehensive evaluation of six distinct tokenization strategies by pretraining transformer-based language models for Nepali and evaluating their performance across multiple downstream tasks. While recent prominent models like GPT, RoBERTa, Claude, LLaMA, Mistral, Falcon, and MPT have adopted byte-level BPE tokenization, our findings demonstrate that for Nepali, SentencePiece tokenization consistently yields superior results on understanding-based tasks. Unlike previous studies that primarily focused on BERT-based architectures, our research specifically examines sequential transformer models, providing valuable insights for language model development in low-resource languages and highlighting the importance of tokenization strategy beyond perplexity reduction. | [
{
"version": "v1",
"created": "Sun, 28 Apr 2024 05:26:12 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 10:43:57 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Luitel",
"Nishant",
""
],
[
"Bekoju",
"Nirajan",
""
],
[
"Sah",
"Anand Kumar",
""
],
[
"Shakya",
"Subarna",
""
]
] |
2404.18155 | Jinming Cao | Jinming Cao, Sicheng Shen, Qiu Zhou, Yifang Yin, Yangyan Li and Roger Zimmermann | ShapeMoir\'e: Channel-Wise Shape-Guided Network for Image Demoir\'eing | 19 pages | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Photographing optoelectronic displays often introduces unwanted moir\'e patterns due to analog signal interference between the pixel grids of the display and the camera sensor arrays. This work identifies two problems that are largely ignored by existing image demoir\'eing approaches: 1) moir\'e patterns vary across different channels (RGB); 2) repetitive patterns are constantly observed. However, employing conventional convolutional (CNN) layers cannot address these problems. Instead, this paper presents the use of our recently proposed \emph{Shape} concept. It was originally employed to model consistent features from fragmented regions, particularly when identical or similar objects coexist in an RGB-D image. Interestingly, we find that the Shape information effectively captures the moir\'e patterns in artifact images. Motivated by this discovery, we propose a new method, ShapeMoir\'e, for image demoir\'eing. Beyond modeling shape features at the patch-level, we further extend this to the global image-level and design a novel Shape-Architecture. Consequently, our proposed method, equipped with both ShapeConv and Shape-Architecture, can be seamlessly integrated into existing approaches without introducing any additional parameters or computation overhead during inference. We conduct extensive experiments on four widely used datasets, and the results demonstrate that our ShapeMoir\'e achieves state-of-the-art performance, particularly in terms of the PSNR metric. | [
{
"version": "v1",
"created": "Sun, 28 Apr 2024 12:12:08 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 13:33:13 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Cao",
"Jinming",
""
],
[
"Shen",
"Sicheng",
""
],
[
"Zhou",
"Qiu",
""
],
[
"Yin",
"Yifang",
""
],
[
"Li",
"Yangyan",
""
],
[
"Zimmermann",
"Roger",
""
]
] |
2404.18329 | Paolo Fachin | Paolo Fachin, Francesco Macheda, Paolo Barone, Francesco Mauri | Quantized Born Effective charges as probes for the topological phase transition in the Haldane and Kane-Mele models | null | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | We propose a new approach to study the transition between different topological states, based on the assessment of the vibrational resonances in infrared spectra. We consider the Haldane and Kane-Mele models finding that Born effective charges are nearly quantized, with a discontinuous jump concomitant with the topological phase transition. In particular, Born effective charges display a finite value in the trivial phase and a null one in the nontrivial one. This is rooted in the connection between Born effective charges and electronic Berry curvature at the band edges. Finally, at the topological phase transition of the Haldane model, we also observe a nearly quantized jump of the chiral splitting of the zone-center phonon frequencies, induced by time-reversal symmetry breaking. | [
{
"version": "v1",
"created": "Sun, 28 Apr 2024 22:37:29 GMT"
},
{
"version": "v2",
"created": "Tue, 30 Apr 2024 21:59:36 GMT"
},
{
"version": "v3",
"created": "Wed, 13 Nov 2024 10:46:01 GMT"
},
{
"version": "v4",
"created": "Sat, 07 Jun 2025 17:58:04 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Fachin",
"Paolo",
""
],
[
"Macheda",
"Francesco",
""
],
[
"Barone",
"Paolo",
""
],
[
"Mauri",
"Francesco",
""
]
] |
2404.18512 | Peng He | Peng He, Jing-Xin Liu, Hong Wu, Z. D. Wang | Floquet Amorphous Topological Orders in a One-dimensional Rydberg Glass | 8 pages, 4 figures | Communications Physics 8, 237 (2025) | 10.1038/s42005-025-02164-4 | null | cond-mat.mes-hall cond-mat.dis-nn cond-mat.quant-gas | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The topological orders in amorphous systems that lack crystalline symmetry have gained considerable attention recently. Here we propose the Floquet amorphous topological matter, among which the topological orders are explored in experimentally accessible one-dimensional array of randomly pointed Rydberg atoms with periodic driving. The topological properties are comprehensively characterized, considering both the single-particle and many-body perspectives. It is found that the periodic driving leads to rich topological phases of matter. At the single-particle level, we evaluate the real space winding numbers and polarization, revealing robust amorphous topological phases with 0-type and $\pi$-type edge modes. We show a structural disorder induced topological phase transition associated with localization transition in the nonequilibrium system. Remarkably, in the many-body case it is discovered that the amorphous topological order exists in the chain of hardcore bosons, captured by the topological entanglement entropy and the string order. Moreover, feasible experimental probe protocols are also elaborated. | [
{
"version": "v1",
"created": "Mon, 29 Apr 2024 08:57:26 GMT"
},
{
"version": "v2",
"created": "Sun, 08 Jun 2025 04:53:11 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"He",
"Peng",
""
],
[
"Liu",
"Jing-Xin",
""
],
[
"Wu",
"Hong",
""
],
[
"Wang",
"Z. D.",
""
]
] |
2404.18996 | Subrata Samanta | Debtosh Chowdhury, Poulami Mondal, Subrata Samanta | Next-to-Leading Order Unitarity Fits in the Extended Georgi-Machacek Model | 62 pages, 15 figures. Updated figures; clarifications and references added | null | null | null | hep-ph hep-ex | http://creativecommons.org/licenses/by/4.0/ | The Georgi-Machacek (GM) model is a triplet scalar extension of the Standard Model (SM) that preserves custodial symmetry (CS) due to an explicit global $SU(2)_L\otimes SU(2)_R$ symmetry in the scalar potential at tree-level. However, it is also possible to construct a triplet extended scalar sector of the SM without imposing this global symmetry in the potential while still maintaining CS at tree-level. This is referred to as the extended GM (eGM) model. We compute one-loop corrections to all $2\to2$ bosonic scattering amplitudes in both models and place next-to-leading order (NLO) unitarity bounds on the quartic couplings. Further, we derive the bounded-from-below (BFB) conditions on the quartic couplings demanding the stability of the scalar potential in the field subspaces. We show that the 3-field BFB conditions provide a very good approximation of the 13-field BFB conditions for both models and are computationally more efficient. With these theoretical constraints, we perform a global fit of both models to the latest Higgs signal strength results from the $13$ TeV Large Hadron Collider. We observe that the global fit disfavors the regions where $\kappa_V > 1.05$, $\kappa_V < 0.95$, and $\kappa_f > 1.05$, $\kappa_f< 0.92$ at a $95.4\%$ CL for both models. The global fit results demonstrate that NLO unitarity and stability bounds play a significant role in constraining the allowed parameter space of both models. We obtain an upper limit on the absolute values of the scalar quartic couplings to be $1.91\:(2.51)$ in the GM (eGM) model. We find that in both models, the absolute mass differences between the heavy Higgs bosons are less than $410$ GeV if their masses are below $1.1$ TeV. The maximal mass difference among the members of each CS multiplet is less than $210$ GeV in the eGM model. | [
{
"version": "v1",
"created": "Mon, 29 Apr 2024 18:00:01 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Jun 2025 18:02:10 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Chowdhury",
"Debtosh",
""
],
[
"Mondal",
"Poulami",
""
],
[
"Samanta",
"Subrata",
""
]
] |
2404.19431 | Mohammad Javad Ahmadi | Mohammad Javad Ahmadi, Rafael F. Schaefer, H. Vincent Poor | Toward Practical and Theoretical Integration of Sensing and Communication in URA Systems | null | null | null | null | cs.IT math.IT | http://creativecommons.org/licenses/by/4.0/ | This work addresses the problem of integrated sensing and communications (ISAC) involving a massive number of unsourced and uncoordinated users. In the proposed model, known as the unsourced ISAC system (UNISAC), all active communication and sensing users simultaneously share a short frame to transmit their signals without requiring scheduling by the base station or the need to announce their identities. Consequently, the received signal from each user is heavily affected by interference from numerous other users, making it challenging to extract individual transmissions. UNISAC is designed to decode the message sequences from communication users while simultaneously detecting active sensing users and estimating their angles of arrival, regardless of the senders' identities. We derive an approximate achievable result for UNISAC and demonstrate its advantages over conventional methods, such as ALOHA, time-division multiple access, treating interference as noise, and multiple signal classification. Additionally, we propose a practical model that validates the feasibility of the approximate achievable result, showing comparable or even superior performance in scenarios with a small number of users. Through numerical simulations, we demonstrate the effectiveness of both the practical UNISAC model and the approximate achievable result. | [
{
"version": "v1",
"created": "Tue, 30 Apr 2024 10:26:04 GMT"
},
{
"version": "v2",
"created": "Wed, 01 May 2024 05:17:35 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Sep 2024 22:33:49 GMT"
},
{
"version": "v4",
"created": "Sun, 08 Jun 2025 12:51:50 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Ahmadi",
"Mohammad Javad",
""
],
[
"Schaefer",
"Rafael F.",
""
],
[
"Poor",
"H. Vincent",
""
]
] |
2404.19440 | Roman Kuzian | E. E. Krasovskii and R.O. Kuzian | Negative transit time in non-tunneling electron transmission through graphene multilayers | 9 pages, 6 figures | null | 10.1103/PhysRevLett.134.126302 | null | cond-mat.mtrl-sci physics.comp-ph physics.optics quant-ph | http://creativecommons.org/licenses/by/4.0/ | Attosecond dynamics of electron transmission through atomically-thin crystalline films is studied with an {\em ab initio} scattering theory. The temporal character of the electron propagation through graphene multilayers is traced to the band structure of bulk graphite: In the forbidden gaps the wave packet transit time $\tau_\mathrm{T}$ saturates with thickness and in the allowed bands $\tau_\mathrm{T}$ oscillates following transmission resonances. Hitherto unknown negative transit time due to in-plane scattering is discovered in monolayers of graphene, h-BN, and oxygen. Moreover, Wigner time delay is found to diverge at the scattering resonances caused by the emergence of secondary diffracted beams. This offers a way to manipulate the propagation timing of the wave packet without sacrificing the transmitted intensity. The spatial reshaping of the wave packet at the resonances may help elucidate details of the streaking by an inhomogeneous field at the surface. | [
{
"version": "v1",
"created": "Tue, 30 Apr 2024 10:43:04 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Krasovskii",
"E. E.",
""
],
[
"Kuzian",
"R. O.",
""
]
] |
2405.00228 | Hatef Otroshi Shahreza | David Geissb\"uhler, Hatef Otroshi Shahreza, and S\'ebastien Marcel | Synthetic Face Datasets Generation via Latent Space Exploration from Brownian Identity Diffusion | Accepted in ICML 2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Face recognition models are trained on large-scale datasets, which have privacy and ethical concerns. Lately, the use of synthetic data to complement or replace genuine data for the training of face recognition models has been proposed. While promising results have been obtained, it still remains unclear if generative models can yield diverse enough data for such tasks. In this work, we introduce a new method, inspired by the physical motion of soft particles subjected to stochastic Brownian forces, allowing us to sample identities distributions in a latent space under various constraints. We introduce three complementary algorithms, called Langevin, Dispersion, and DisCo, aimed at generating large synthetic face datasets. With this in hands, we generate several face datasets and benchmark them by training face recognition models, showing that data generated with our method exceeds the performance of previously GAN-based datasets and achieves competitive performance with state-of-the-art diffusion-based synthetic datasets. While diffusion models are shown to memorize training data, we prevent leakage in our new synthetic datasets, paving the way for more responsible synthetic datasets. | [
{
"version": "v1",
"created": "Tue, 30 Apr 2024 22:32:02 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Jun 2025 20:01:04 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Geissbühler",
"David",
""
],
[
"Shahreza",
"Hatef Otroshi",
""
],
[
"Marcel",
"Sébastien",
""
]
] |
2405.01627 | Haoyu Guo | Haoyu Guo, Rohit Mukherjee and Debanjan Chowdhury | Dynamical Freezing in Exactly Solvable Models of Driven Chaotic Quantum Dots | (v1) 5 pages, 3 figures with supplement; (v2) 5 pages, 3 figures with end matter and supplement. Added clarification to main text. Added discussion for the time-dependent single-particle spectrum in the supplement; (v4) updated references | Phys. Rev. Lett. 134, 226501 (2025) | 10.1103/ggk3-6cf8 | null | cond-mat.str-el cond-mat.mes-hall cond-mat.stat-mech hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The late-time equilibrium behavior of generic interacting models is determined by the coupled hydrodynamic equations associated with the globally conserved quantities. In the presence of an external time-dependent drive, non-integrable systems typically thermalize to an effectively infinite-temperature state, losing all memory of their initial states. However, in the presence of a large time-periodic Floquet drive, there exist special points in phase-space where the strongly interacting system develops approximate {\it emergent} conservation laws. Here we present results for an exactly solvable model of two coupled chaotic quantum dots with multiple orbitals interacting via random two and four-fermion interactions in the presence of a Floquet drive. We analyze the phenomenology of dynamically generated freezing using a combination of exact diagonalization, and field-theoretic analysis in the limit of a large number of electronic orbitals. The model displays universal freezing behavior irrespective of whether the theory is averaged over the disorder configurations or not. We present explicit computations for the growth of many-body chaos and entanglement entropy, which demonstrates the long-lived coherence associated with the interacting degrees of freedom even at late-times at the dynamically frozen points. We also compute the slow timescale that controls relaxation away from exact freezing in a high-frequency expansion. | [
{
"version": "v1",
"created": "Thu, 02 May 2024 18:00:01 GMT"
},
{
"version": "v2",
"created": "Sat, 28 Sep 2024 19:16:11 GMT"
},
{
"version": "v3",
"created": "Thu, 08 May 2025 03:16:51 GMT"
},
{
"version": "v4",
"created": "Sat, 07 Jun 2025 17:52:11 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Guo",
"Haoyu",
""
],
[
"Mukherjee",
"Rohit",
""
],
[
"Chowdhury",
"Debanjan",
""
]
] |
2405.01786 | Byeongseon Go | Byeongseon Go, Changhun Oh, and Hyunseok Jeong | On computational complexity and average-case hardness of shallow-depth boson sampling | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Boson sampling, a computational task believed to be classically hard to simulate, is expected to hold promise for demonstrating quantum computational advantage using near-term quantum devices. However, noise in experimental implementations poses a significant challenge, potentially rendering boson sampling classically simulable and compromising its classical intractability. Numerous studies have proposed classical algorithms under various noise models that can efficiently simulate boson sampling as noise rates increase with circuit depth. To address this issue particularly related to circuit depth, we explore the viability of achieving quantum computational advantage through boson sampling with shallow-depth linear optical circuits. Specifically, as the average-case hardness of estimating output probabilities of boson sampling is a crucial ingredient in demonstrating its classical intractability, we make progress on establishing the average-case hardness confined to logarithmic-depth regimes. We also obtain the average-case hardness for logarithmic-depth Fock-state boson sampling subject to lossy environments and for the logarithmic-depth Gaussian boson sampling. By providing complexity-theoretical backgrounds for the classical simulation hardness of logarithmic-depth boson sampling, we expect that our findings will mark a crucial step towards a more noise-tolerant demonstration of quantum advantage with shallow-depth boson sampling. | [
{
"version": "v1",
"created": "Fri, 03 May 2024 00:12:48 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Jun 2025 03:29:41 GMT"
}
] | 2025-06-10T00:00:00 | [
[
"Go",
"Byeongseon",
""
],
[
"Oh",
"Changhun",
""
],
[
"Jeong",
"Hyunseok",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.