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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.02460 | Yupeng Qi | Yupeng Qi, Ziyu Lyu, Min Yang, Yanlin Wang, Lu Bai, Lixin Cui | MidPO: Dual Preference Optimization for Safety and Helpfulness in Large Language Models via a Mixture of Experts Framework | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As large language models (LLMs) are increasingly applied across various domains, enhancing safety while maintaining the helpfulness of LLMs has become a critical challenge. Recent studies solve this problem through safety-constrained online preference optimization or safety-constrained offline preference optimization. However, the safety-constrained online methods often suffer from excessive safety, which might reduce helpfulness, while the safety-constrained offline methods perform poorly in adaptively balancing safety and helpfulness. To address these limitations, we propose MidPO, a \textbf{\underline{Mi}}xture of Experts (MoE) framework for safety-helpfulness \textbf{\underline{d}}ual \textbf{\underline{P}}reference \textbf{\underline{O}}ptimization. Firstly, MidPO devises single-preference enhanced direct preference optimization approach to transform the base model into two independent experts, termed safety and helpfulness experts, and fine-tunes the two independent experts for optimal safety or helpfulness performance. Secondly, to achieve an effective balance between safety and helpfulness, MidPO incorporates the two experts into the MoE framework and designs a dynamic routing mechanism to allocate contributions from each expert adaptively. We conduct quantitative and qualitative experiments on three popular datasets to demonstrate the proposed MidPO significantly outperforms state-of-the-art approaches in both safety and helpfulness. The code and models will be released. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:23:09 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Qi",
"Yupeng",
""
],
[
"Lyu",
"Ziyu",
""
],
[
"Yang",
"Min",
""
],
[
"Wang",
"Yanlin",
""
],
[
"Bai",
"Lu",
""
],
[
"Cui",
"Lixin",
""
]
] |
2506.02461 | Chunkit Chan | Chunkit Chan, Yauwai Yim, Hongchuan Zeng, Zhiying Zou, Xinyuan Cheng, Zhifan Sun, Zheye Deng, Kawai Chung, Yuzhuo Ao, Yixiang Fan, Cheng Jiayang, Ercong Nie, Ginny Y. Wong, Helmut Schmid, Hinrich Sch\"utze, Simon See, Yangqiu Song | XToM: Exploring the Multilingual Theory of Mind for Large Language Models | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Theory of Mind (ToM), the ability to infer mental states in others, is pivotal for human social cognition. Existing evaluations of ToM in LLMs are largely limited to English, neglecting the linguistic diversity that shapes human cognition. This limitation raises a critical question: can LLMs exhibit Multilingual Theory of Mind, which is the capacity to reason about mental states across diverse linguistic contexts? To address this gap, we present XToM, a rigorously validated multilingual benchmark that evaluates ToM across five languages and incorporates diverse, contextually rich task scenarios. Using XToM, we systematically evaluate LLMs (e.g., DeepSeek R1), revealing a pronounced dissonance: while models excel in multilingual language understanding, their ToM performance varies across languages. Our findings expose limitations in LLMs' ability to replicate human-like mentalizing across linguistic contexts. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:23:25 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Chan",
"Chunkit",
""
],
[
"Yim",
"Yauwai",
""
],
[
"Zeng",
"Hongchuan",
""
],
[
"Zou",
"Zhiying",
""
],
[
"Cheng",
"Xinyuan",
""
],
[
"Sun",
"Zhifan",
""
],
[
"Deng",
"Zheye",
""
],
[
"Chung",
"Kawai",
""
],
[
"Ao",
"Yuzhuo",
""
],
[
"Fan",
"Yixiang",
""
],
[
"Jiayang",
"Cheng",
""
],
[
"Nie",
"Ercong",
""
],
[
"Wong",
"Ginny Y.",
""
],
[
"Schmid",
"Helmut",
""
],
[
"Schütze",
"Hinrich",
""
],
[
"See",
"Simon",
""
],
[
"Song",
"Yangqiu",
""
]
] |
2506.02462 | Kunyu Wang | Kunyu Wang, Xueyang Fu, Xin Lu, Chengjie Ge, Chengzhi Cao, Wei Zhai, Zheng-Jun Zha | Efficient Test-time Adaptive Object Detection via Sensitivity-Guided Pruning | Accepted as CVPR 2025 oral paper | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Continual test-time adaptive object detection (CTTA-OD) aims to online adapt a source pre-trained detector to ever-changing environments during inference under continuous domain shifts. Most existing CTTA-OD methods prioritize effectiveness while overlooking computational efficiency, which is crucial for resource-constrained scenarios. In this paper, we propose an efficient CTTA-OD method via pruning. Our motivation stems from the observation that not all learned source features are beneficial; certain domain-sensitive feature channels can adversely affect target domain performance. Inspired by this, we introduce a sensitivity-guided channel pruning strategy that quantifies each channel based on its sensitivity to domain discrepancies at both image and instance levels. We apply weighted sparsity regularization to selectively suppress and prune these sensitive channels, focusing adaptation efforts on invariant ones. Additionally, we introduce a stochastic channel reactivation mechanism to restore pruned channels, enabling recovery of potentially useful features and mitigating the risks of early pruning. Extensive experiments on three benchmarks show that our method achieves superior adaptation performance while reducing computational overhead by 12% in FLOPs compared to the recent SOTA method. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:27:56 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Wang",
"Kunyu",
""
],
[
"Fu",
"Xueyang",
""
],
[
"Lu",
"Xin",
""
],
[
"Ge",
"Chengjie",
""
],
[
"Cao",
"Chengzhi",
""
],
[
"Zhai",
"Wei",
""
],
[
"Zha",
"Zheng-Jun",
""
]
] |
2506.02463 | Fizaan Khan | Fizaan Khan (1), Sachin Verma (1), Bishwanath Bhoi (1), Rajeev Singh (1) ((1) Nano-Magnetism and Quantum Technology Lab, Department of Physics, Indian Institute of Technology (BHU), Varanasi, India) | Controlling photon-magnon coupling using another magnon | 6 pages, 4 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We investigate photon-mediated magnon-magnon coupling between spatially separated Yttrium Iron Garnet (YIG) and permalloy (NiFe) thin films placed on a planar hexagonal ring resonator. We observe clear signatures of magnon-magnon interaction in the absence of direct dipolar interaction between the magnetic films. Notably, the coupling strength between the hexagonal ring resonator and the permalloy film increases with the thickness of the YIG film, despite a fixed permalloy film thickness, suggesting the presence of an indirect interaction channel mediated by resonator photons. To support these findings, we present a theoretical model that accurately reproduces the observed transmission spectra and reveals a nontrivial interdependence between the individual coupling strengths of YIG and permalloy to the resonator. These results underscore the importance of indirect interactions and potential crosstalk pathways in designing hybrid magnonic systems and scalable quantum architectures, while demonstrating the feasibility of cost-effective, planar configurations for experimental implementation. These insights are valuable for advancing low-loss, coherent information transfer in hybrid quantum devices. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:28:45 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Khan",
"Fizaan",
"",
"Nano-Magnetism and Quantum Technology Lab, Department of Physics, Indian Institute of Technology"
],
[
"Verma",
"Sachin",
"",
"Nano-Magnetism and Quantum Technology Lab, Department of Physics, Indian Institute of Technology"
],
[
"Bhoi",
"Bishwanath",
"",
"Nano-Magnetism and Quantum Technology Lab, Department of Physics, Indian Institute of Technology"
],
[
"Singh",
"Rajeev",
"",
"Nano-Magnetism and Quantum Technology Lab, Department of Physics, Indian Institute of Technology"
]
] |
2506.02464 | Harleen Dahiya Dr. | Navpreet Kaur and Harleen Dahiya | Chiral-odd generalized parton distributions for the low-lying octet baryons | 24 pages, 7 figures and 1 table | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We have studied the chiral-odd generalized parton distributions (GPDs) for the octet baryons within the framework of diquark spectator model, emphasizing the difference arising from different quark flavors of $p$, $\Sigma^+$ and $\Xi^o$. The quark-quark correlator for the tensor current has been solved to investigate the transversely polarized quark dynamics using light-cone wave functions. In the forward limit, transversity distributions of all the constituent quark flavors of $p$, $\Sigma^+$, and $\Xi^o$ have been studied. Additionally, the lowest moment of GPD $H_T$ and the combination of GPDs $E_T$ and $\tilde{H}_T$ have also been investigated. Our results are comparable with other available model predictions and lattice data for the case of $p$. Furthermore, we have presented a comparative analysis of tensor charges and anomalous tensor magnetic moments for all the constituent quark flavors of the considered baryons, comparing them with available data obtained from different approaches. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:31:31 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kaur",
"Navpreet",
""
],
[
"Dahiya",
"Harleen",
""
]
] |
2506.02465 | Lingfeng Chen | Lingfeng Chen, Panhe Hu, Zhiliang Pan, Qi Liu, and Zhen Liu | Large Language Models Can Achieve Explainable and Training-Free One-shot HRRP ATR | Submitted to IEEE SPL 2025 | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This letter introduces a pioneering, training-free and explainable framework for High-Resolution Range Profile (HRRP) automatic target recognition (ATR) utilizing large-scale pre-trained Large Language Models (LLMs). Diverging from conventional methods requiring extensive task-specific training or fine-tuning, our approach converts one-dimensional HRRP signals into textual scattering center representations. Prompts are designed to align LLMs' semantic space for ATR via few-shot in-context learning, effectively leveraging its vast pre-existing knowledge without any parameter update. We make our codes publicly available to foster research into LLMs for HRRP ATR. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:32:17 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Chen",
"Lingfeng",
""
],
[
"Hu",
"Panhe",
""
],
[
"Pan",
"Zhiliang",
""
],
[
"Liu",
"Qi",
""
],
[
"Liu",
"Zhen",
""
]
] |
2506.02466 | Xue Han | Xue Han, Jason Qu, Hengxin Tan, Zicheng Tao, Noah M. Meyer, Patrick S. Kirchmann, Yanfeng Guo, Binghai Yan, Zhi-Xun Shen, Jonathan A. Sobota | Reconstructing the wavefunction of magnetic topological insulators MnBi2Te4 and MnBi4Te7 using spin-resolved photoemission | null | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Despite their importance for exotic quantum effects, the surface electronic structure of magnetic topological insulators MnBi2Te4 and MnBi4Te7 remains poorly understood. Using high-efficiency spin- and angle-resolved photoemission spectroscopy, we directly image the spin-polarization and orbital character of the surface states in both compounds and map our observations onto a model wavefunction to describe the complex spin-orbital texture, which solidifies our understanding of the surface band structure by establishing the single-band nature of the most prominent states. Most importantly, our analysis reveals a new mechanism for reducing the magnetic gap of the topological surface states based on the orbital composition of the wavefunction. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:32:30 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Han",
"Xue",
""
],
[
"Qu",
"Jason",
""
],
[
"Tan",
"Hengxin",
""
],
[
"Tao",
"Zicheng",
""
],
[
"Meyer",
"Noah M.",
""
],
[
"Kirchmann",
"Patrick S.",
""
],
[
"Guo",
"Yanfeng",
""
],
[
"Yan",
"Binghai",
""
],
[
"Shen",
"Zhi-Xun",
""
],
[
"Sobota",
"Jonathan A.",
""
]
] |
2506.02467 | Haowen Pang | Haowen Pang, Weiyan Guo, Chuyang Ye | Multi-modal brain MRI synthesis based on SwinUNETR | 9 pages, 5 figures | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-modal brain magnetic resonance imaging (MRI) plays a crucial role in clinical diagnostics by providing complementary information across different imaging modalities. However, a common challenge in clinical practice is missing MRI modalities. In this paper, we apply SwinUNETR to the synthesize of missing modalities in brain MRI. SwinUNETR is a novel neural network architecture designed for medical image analysis, integrating the strengths of Swin Transformer and convolutional neural networks (CNNs). The Swin Transformer, a variant of the Vision Transformer (ViT), incorporates hierarchical feature extraction and window-based self-attention mechanisms, enabling it to capture both local and global contextual information effectively. By combining the Swin Transformer with CNNs, SwinUNETR merges global context awareness with detailed spatial resolution. This hybrid approach addresses the challenges posed by the varying modality characteristics and complex brain structures, facilitating the generation of accurate and realistic synthetic images. We evaluate the performance of SwinUNETR on brain MRI datasets and demonstrate its superior capability in generating clinically valuable images. Our results show significant improvements in image quality, anatomical consistency, and diagnostic value. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:33:52 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Pang",
"Haowen",
""
],
[
"Guo",
"Weiyan",
""
],
[
"Ye",
"Chuyang",
""
]
] |
2506.02468 | Sathish Kumar | Puja Sonawane and A. Sathish Kumar | On Hermite type sampling Kantorovich operators in the settings of mixed norm Spaces | null | null | null | null | math.FA | http://creativecommons.org/licenses/by/4.0/ | In this paper, we analyze the convergence behavior of Hermite-type sampling Kantorovich operators in the context of mixed norm spaces. We prove certain direct approximation theorems, including the uniform convergence theorem, the Voronovskaja-type asymptotic formula, and an estimate of error in the approximation in terms of the modulus of continuity in mixed norm settings. Next, we estimate the rate of convergence of these sampling Kantorovich operators in terms of the modulus of continuity. In addition, we obtain simultaneous approximation results of these sampling Kantorovich operators, including the uniform approximation, the asymptotic formula, and the approximation error in terms of the modulus of continuity in mixed settings. Finally, using cardinal $B$-splines, the implementation of differentiable functions has been shown. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:34:17 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Sonawane",
"Puja",
""
],
[
"Kumar",
"A. Sathish",
""
]
] |
2506.02469 | Andrea Fagnani | Andrea Fagnani, Bernd Helber, Damien Le Quang, Alessandro Turchi, Jimmy Freitas Monteiro, Annick Hubin, Olivier Chazot | Characterization of the VKI Plasmatron subsonic ICP jet combining optical emission spectroscopy, intrusive measurements, and CFD simulations | null | null | null | null | physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | This paper addresses the characterization of the subsonic flow in the 1.2~MW Inductively Coupled Plasma (ICP) wind tunnel at the von Karman Institute for Fluid Dynamics (VKI), targeting chamber pressures of 50 and 100~mbar, and input electric powers between 150 and 300~kW. Ultraviolet to near-infrared optical emission spectroscopy measurements of the free-jet flow are carried out with an updated experimental set-up, calibration procedure, and data processing, providing high-quality absolute spatially-resolved emission spectra. Emission measurements agree with thermochemical equilibrium predictions within a range of conditions, allowing to extract experimental maps of cold-wall heat flux and dynamic pressure against the inferred free-jet enthalpy. A detailed comparison with the characterization methodology traditionally employed is presented, highlighting the need for an improved modeling strategy. Using the measured free-jet temperature and dynamic pressure only, a forward procedure for the computation of the stagnation line flow is proposed. The latter agrees with intrusive heat flux measurements through a range of test conditions, and for values of the recombination coefficient of the reference copper probe commonly found in the literature. Results demonstrate that a consistent framework between numerical simulations and experimental data can be achieved, defining an improved framework for the characterization of the subsonic ICP jet. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:36:18 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Fagnani",
"Andrea",
""
],
[
"Helber",
"Bernd",
""
],
[
"Quang",
"Damien Le",
""
],
[
"Turchi",
"Alessandro",
""
],
[
"Monteiro",
"Jimmy Freitas",
""
],
[
"Hubin",
"Annick",
""
],
[
"Chazot",
"Olivier",
""
]
] |
2506.02470 | Xuejiao Zhao | Xuejiao Zhao, Siyan Liu, Su-Yin Yang, Chunyan Miao | A Smart Multimodal Healthcare Copilot with Powerful LLM Reasoning | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Misdiagnosis causes significant harm to healthcare systems worldwide, leading to increased costs and patient risks. MedRAG is a smart multimodal healthcare copilot equipped with powerful large language model (LLM) reasoning, designed to enhance medical decision-making. It supports multiple input modalities, including non-intrusive voice monitoring, general medical queries, and electronic health records. MedRAG provides recommendations on diagnosis, treatment, medication, and follow-up questioning. Leveraging retrieval-augmented generation enhanced by knowledge graph-elicited reasoning, MedRAG retrieves and integrates critical diagnostic insights, reducing the risk of misdiagnosis. It has been evaluated on both public and private datasets, outperforming existing models and offering more specific and accurate healthcare assistance. A demonstration video of MedRAG is available at: https://www.youtube.com/watch?v=PNIBDMYRfDM. The source code is available at: https://github.com/SNOWTEAM2023/MedRAG. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:39:02 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zhao",
"Xuejiao",
""
],
[
"Liu",
"Siyan",
""
],
[
"Yang",
"Su-Yin",
""
],
[
"Miao",
"Chunyan",
""
]
] |
2506.02471 | Bauyrzhan Sartayev | B. Sartayev | 4-type subvarieties of the variety of associative algebras | 10 pages | null | null | null | math.RA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we consider four subvarieties of the variety of associative algebras. We study these subvarieties from the point of view of operads and show connections with well-known algebras such as dendriform algebras and noncommutative Novikov algebras. Finally, we define operations, commutator and anti-commutator on these algebras and obtain some identities under these operations. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:40:50 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Sartayev",
"B.",
""
]
] |
2506.02472 | Halil Helvaci | Halil Ismail Helvaci, Justin Philip Huber, Jihye Bae, Sen-ching Samson Cheung | HRTR: A Single-stage Transformer for Fine-grained Sub-second Action Segmentation in Stroke Rehabilitation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Stroke rehabilitation often demands precise tracking of patient movements to monitor progress, with complexities of rehabilitation exercises presenting two critical challenges: fine-grained and sub-second (under one-second) action detection. In this work, we propose the High Resolution Temporal Transformer (HRTR), to time-localize and classify high-resolution (fine-grained), sub-second actions in a single-stage transformer, eliminating the need for multi-stage methods and post-processing. Without any refinements, HRTR outperforms state-of-the-art systems on both stroke related and general datasets, achieving Edit Score (ES) of 70.1 on StrokeRehab Video, 69.4 on StrokeRehab IMU, and 88.4 on 50Salads. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:42:15 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Helvaci",
"Halil Ismail",
""
],
[
"Huber",
"Justin Philip",
""
],
[
"Bae",
"Jihye",
""
],
[
"Cheung",
"Sen-ching Samson",
""
]
] |
2506.02473 | Xinran (Nicole) Han | Xinran Nicole Han, Ko Nishino, Todd Zickler | Generative Perception of Shape and Material from Differential Motion | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Perceiving the shape and material of an object from a single image is inherently ambiguous, especially when lighting is unknown and unconstrained. Despite this, humans can often disentangle shape and material, and when they are uncertain, they often move their head slightly or rotate the object to help resolve the ambiguities. Inspired by this behavior, we introduce a novel conditional denoising-diffusion model that generates samples of shape-and-material maps from a short video of an object undergoing differential motions. Our parameter-efficient architecture allows training directly in pixel-space, and it generates many disentangled attributes of an object simultaneously. Trained on a modest number of synthetic object-motion videos with supervision on shape and material, the model exhibits compelling emergent behavior: For static observations, it produces diverse, multimodal predictions of plausible shape-and-material maps that capture the inherent ambiguities; and when objects move, the distributions quickly converge to more accurate explanations. The model also produces high-quality shape-and-material estimates for less ambiguous, real-world objects. By moving beyond single-view to continuous motion observations, our work suggests a generative perception approach for improving visual reasoning in physically-embodied systems. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:43:20 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Han",
"Xinran Nicole",
""
],
[
"Nishino",
"Ko",
""
],
[
"Zickler",
"Todd",
""
]
] |
2506.02474 | Keita Nakmaura | Keita Nakamura, Tomoyuki Nakagawa, Kouji Tahata | Quasi-symmetry and geometric marginal homogeneity: A simplicial approach to square contingency tables | null | null | null | null | math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Square contingency tables are traditionally analyzed with a focus on the symmetric structure of the corresponding probability tables. We view probability tables as elements of a simplex equipped with the Aitchison geometry. This perspective allows us to present a novel approach to analyzing symmetric structure using a compositionally coherent framework. We present a geometric interpretation of quasi-symmetry as an e-flat subspace and introduce a new concept called geometric marginal homogeneity, which is also characterized as an e-flat structure. We prove that both quasi-symmetric tables and geometric marginal homogeneous tables form subspaces in the simplex, and demonstrate that the measure of skew-symmetry in Aitchison geometry can be orthogonally decomposed into measures of departure from quasi-symmetry and geometric marginal homogeneity. We illustrate the application and effectiveness of our proposed methodology using data on unaided distance vision from a sample of women. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:44:04 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Nakamura",
"Keita",
""
],
[
"Nakagawa",
"Tomoyuki",
""
],
[
"Tahata",
"Kouji",
""
]
] |
2506.02475 | Jiaxi Hu | Jiaxi Hu, Yongqi Pan, Jusen Du, Disen Lan, Xiaqiang Tang, Qingsong Wen, Yuxuan Liang, Weigao Sun | Comba: Improving Nonlinear RNNs with Closed-loop Control | null | null | null | null | cs.LG cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent efficient sequence modeling methods such as Gated DeltaNet, TTT, and RWKV-7 have achieved performance improvements by supervising the recurrent memory management through Delta learning rule. Unlike previous state-space models (e.g., Mamba) and gated linear attentions (e.g., GLA), these models introduce interactions between the recurrent state and the key vector, resulting in a nonlinear recursive structure. In this paper, we first introduce the concept of Nonlinear RNNs with a comprehensive analysis on the advantages and limitations of these models. Then, based on closed-loop control theory, we propose a novel Nonlinear RNN variant named Comba, which adopts a scalar-plus-low-rank state transition, with both state feedback and output feedback corrections. We also implement a hardware-efficient chunk-wise parallel kernel in Triton and train models with 340M/1.3B parameters on large-scale corpus. Comba demonstrates its superior performance and computation efficiency in both language and vision modeling. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:44:50 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Hu",
"Jiaxi",
""
],
[
"Pan",
"Yongqi",
""
],
[
"Du",
"Jusen",
""
],
[
"Lan",
"Disen",
""
],
[
"Tang",
"Xiaqiang",
""
],
[
"Wen",
"Qingsong",
""
],
[
"Liang",
"Yuxuan",
""
],
[
"Sun",
"Weigao",
""
]
] |
2506.02476 | Alessio Liberatori Mr. | Alessio Liberatori, Deimer Antonio Alvarez Garay, Marco Palla, Alessio Mucciarelli, Michele Bellazzini, Donatella Romano | A chemical close-up of the main body of the Sagittarius dwarf galaxy | Accepted in A&A | null | null | null | astro-ph.GA astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present the chemical composition of a sample of 37 red giant branch (RGB) stars belonging to the main body of the remnant of the Sagittarius (Sgr) dwarf spheroidal galaxy. All stars were observed with the FLAMES-UVES high-resolution spectrograph. Twenty-three new targets are selected along the blue side of the RGB of Sgr, but outside the galaxy stellar nucleus, in order to avoid contamination by the stars of the metal-poor globular cluster M54. Additionally, we re-analyzed archival spectra of fourteen targets located on the red RGB. For this sample, we derive the abundances of 21 chemical species (from Oxygen to Europium) representing different nucleosynthetic sites. The sample covers a large range of metallicity, from [Fe/H]~-2 to ~ -0.4 dex and we can identify the transition between the enrichment phases dominated by core-collapse (CC-SNe) and Type Ia (SNe-Ia) supernovae. The observed [{\alpha}/Fe] trend suggests a knee occurring at [Fe/H]~-1.5/-1.3 dex, compatible with the rather low star formation efficiency of Sgr. At lower [Fe/H], Sgr stars exhibit a chemical composition compatible with Milky Way stars of similar [Fe/H]. The only relevant exceptions are [Mn/Fe], [Zn/Fe], and [Eu/Fe]. At [Fe/H] higher than ~ -1.5/-1.3 dex, instead, the chemical pattern of Sgr significantly deviates from that of the Milky Way for almost all the elements analyzed in this study. Some of the abundance patterns reveal a lower contribution by very massive stars exploding as hypernovae (e.g. [Mn/Fe], [Zn/Fe]), a higher contribution by sub-Chandrasekhar progenitors of SNe Ia (e.g. [Ni/Fe]) and a high production efficiency of rapid neutron-capture elements ([Eu/Fe]). | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:46:46 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Liberatori",
"Alessio",
""
],
[
"Garay",
"Deimer Antonio Alvarez",
""
],
[
"Palla",
"Marco",
""
],
[
"Mucciarelli",
"Alessio",
""
],
[
"Bellazzini",
"Michele",
""
],
[
"Romano",
"Donatella",
""
]
] |
2506.02477 | Kunyu Wang | Kunyu Wang, Xueyang Fu, Chengzhi Cao, Chengjie Ge, Wei Zhai, Zheng-Jun Zha | Towards Better De-raining Generalization via Rainy Characteristics Memorization and Replay | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Current image de-raining methods primarily learn from a limited dataset, leading to inadequate performance in varied real-world rainy conditions. To tackle this, we introduce a new framework that enables networks to progressively expand their de-raining knowledge base by tapping into a growing pool of datasets, significantly boosting their adaptability. Drawing inspiration from the human brain's ability to continuously absorb and generalize from ongoing experiences, our approach borrow the mechanism of the complementary learning system. Specifically, we first deploy Generative Adversarial Networks (GANs) to capture and retain the unique features of new data, mirroring the hippocampus's role in learning and memory. Then, the de-raining network is trained with both existing and GAN-synthesized data, mimicking the process of hippocampal replay and interleaved learning. Furthermore, we employ knowledge distillation with the replayed data to replicate the synergy between the neocortex's activity patterns triggered by hippocampal replays and the pre-existing neocortical knowledge. This comprehensive framework empowers the de-raining network to amass knowledge from various datasets, continually enhancing its performance on previously unseen rainy scenes. Our testing on three benchmark de-raining networks confirms the framework's effectiveness. It not only facilitates continuous knowledge accumulation across six datasets but also surpasses state-of-the-art methods in generalizing to new real-world scenarios. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:50:00 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Wang",
"Kunyu",
""
],
[
"Fu",
"Xueyang",
""
],
[
"Cao",
"Chengzhi",
""
],
[
"Ge",
"Chengjie",
""
],
[
"Zhai",
"Wei",
""
],
[
"Zha",
"Zheng-Jun",
""
]
] |
2506.02478 | Zijian Li | Zijian Li, Xiaocheng Feng, Huixin Liu, Yichong Huang, Ting Liu, Bing Qin | FroM: Frobenius Norm-Based Data-Free Adaptive Model Merging | 12 pages, 11 figures | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the development of large language models, fine-tuning has emerged as an effective method to enhance performance in specific scenarios by injecting domain-specific knowledge. In this context, model merging techniques provide a solution for fusing knowledge from multiple fine-tuning models by combining their parameters. However, traditional methods often encounter task interference when merging full fine-tuning models, and this problem becomes even more evident in parameter-efficient fine-tuning scenarios. In this paper, we introduce an improvement to the RegMean method, which indirectly leverages the training data to approximate the outputs of the linear layers before and after merging. We propose an adaptive merging method called FroM, which directly measures the model parameters using the Frobenius norm, without any training data. By introducing an additional hyperparameter for control, FroM outperforms baseline methods across various fine-tuning scenarios, alleviating the task interference problem. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:50:09 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Li",
"Zijian",
""
],
[
"Feng",
"Xiaocheng",
""
],
[
"Liu",
"Huixin",
""
],
[
"Huang",
"Yichong",
""
],
[
"Liu",
"Ting",
""
],
[
"Qin",
"Bing",
""
]
] |
2506.02479 | Kalyan Nakka | Kalyan Nakka, Nitesh Saxena | BitBypass: A New Direction in Jailbreaking Aligned Large Language Models with Bitstream Camouflage | 24 pages, 24 figures, and 7 tables | null | null | null | cs.CR cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | The inherent risk of generating harmful and unsafe content by Large Language Models (LLMs), has highlighted the need for their safety alignment. Various techniques like supervised fine-tuning, reinforcement learning from human feedback, and red-teaming were developed for ensuring the safety alignment of LLMs. However, the robustness of these aligned LLMs is always challenged by adversarial attacks that exploit unexplored and underlying vulnerabilities of the safety alignment. In this paper, we develop a novel black-box jailbreak attack, called BitBypass, that leverages hyphen-separated bitstream camouflage for jailbreaking aligned LLMs. This represents a new direction in jailbreaking by exploiting fundamental information representation of data as continuous bits, rather than leveraging prompt engineering or adversarial manipulations. Our evaluation of five state-of-the-art LLMs, namely GPT-4o, Gemini 1.5, Claude 3.5, Llama 3.1, and Mixtral, in adversarial perspective, revealed the capabilities of BitBypass in bypassing their safety alignment and tricking them into generating harmful and unsafe content. Further, we observed that BitBypass outperforms several state-of-the-art jailbreak attacks in terms of stealthiness and attack success. Overall, these results highlights the effectiveness and efficiency of BitBypass in jailbreaking these state-of-the-art LLMs. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:51:18 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Nakka",
"Kalyan",
""
],
[
"Saxena",
"Nitesh",
""
]
] |
2506.02480 | Yifan Duan | Yifan Duan, Yihong Tang, Kehai Chen, Liqiang Nie, Min Zhang | ORPP: Self-Optimizing Role-playing Prompts to Enhance Language Model Capabilities | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High-quality prompts are crucial for eliciting outstanding performance from large language models (LLMs) on complex tasks. Existing research has explored model-driven strategies for prompt optimization. However, these methods often suffer from high computational overhead or require strong optimization capabilities from the model itself, which limits their broad applicability.To address these challenges, we propose ORPP (Optimized Role-Playing Prompt),a framework that enhances model performance by optimizing and generating role-playing prompts. The core idea of ORPP is to confine the prompt search space to role-playing scenarios, thereby fully activating the model's intrinsic capabilities through carefully crafted, high-quality role-playing prompts. Specifically, ORPP first performs iterative optimization on a small subset of training samples to generate high-quality role-playing prompts. Then, leveraging the model's few-shot learning capability, it transfers the optimization experience to efficiently generate suitable prompts for the remaining samples.Our experimental results show that ORPP not only matches but in most cases surpasses existing mainstream prompt optimization methods in terms of performance. Notably, ORPP demonstrates superior "plug-and-play" capability. In most cases, it can be integrated with various other prompt methods and further enhance their effectiveness. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:51:35 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Duan",
"Yifan",
""
],
[
"Tang",
"Yihong",
""
],
[
"Chen",
"Kehai",
""
],
[
"Nie",
"Liqiang",
""
],
[
"Zhang",
"Min",
""
]
] |
2506.02481 | Inderjeet Nair | Inderjeet Nair, Lu Wang | Do Language Models Think Consistently? A Study of Value Preferences Across Varying Response Lengths | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Evaluations of LLMs' ethical risks and value inclinations often rely on short-form surveys and psychometric tests, yet real-world use involves long-form, open-ended responses -- leaving value-related risks and preferences in practical settings largely underexplored. In this work, we ask: Do value preferences inferred from short-form tests align with those expressed in long-form outputs? To address this question, we compare value preferences elicited from short-form reactions and long-form responses, varying the number of arguments in the latter to capture users' differing verbosity preferences. Analyzing five LLMs (llama3-8b, gemma2-9b, mistral-7b, qwen2-7b, and olmo-7b), we find (1) a weak correlation between value preferences inferred from short-form and long-form responses across varying argument counts, and (2) similarly weak correlation between preferences derived from any two distinct long-form generation settings. (3) Alignment yields only modest gains in the consistency of value expression. Further, we examine how long-form generation attributes relate to value preferences, finding that argument specificity negatively correlates with preference strength, while representation across scenarios shows a positive correlation. Our findings underscore the need for more robust methods to ensure consistent value expression across diverse applications. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:52:03 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Nair",
"Inderjeet",
""
],
[
"Wang",
"Lu",
""
]
] |
2506.02482 | Liu Minghao Mark | Minghao Liu, Catherine Zhao, Nathan Zhou | Building a Recommendation System Using Amazon Product Co-Purchasing Network | null | null | null | null | cs.SI | http://creativecommons.org/licenses/by/4.0/ | This project develops an online, inductive recommendation system for newly listed products on e-commerce platforms, focusing on suggesting relevant new items to customers as they purchase other products. Using the Amazon Product Co-Purchasing Network Metadata dataset, we construct a co-purchasing graph where nodes represent products and edges capture co-purchasing relationships. To address the challenge of recommending new products with limited information, we apply a modified GraphSAGE method for link prediction. This inductive approach leverages both product features and the existing co-purchasing graph structure to predict potential co-purchasing relationships, enabling the model to generalize to unseen products. As an online method, it updates in real time, making it scalable and adaptive to evolving product catalogs. Experimental results demonstrate that our approach outperforms baseline algorithms in predicting relevant product links, offering a promising solution for enhancing the relevance of new product recommendations in e-commerce environments. All code is available at https://github.com/cse416a-fl24/final-project-l-minghao_z-catherine_z-nathan.git. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:52:22 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Liu",
"Minghao",
""
],
[
"Zhao",
"Catherine",
""
],
[
"Zhou",
"Nathan",
""
]
] |
2506.02483 | Sina Bagheri Nezhad | Sina Bagheri Nezhad and Ameeta Agrawal | Enhancing Large Language Models with Neurosymbolic Reasoning for Multilingual Tasks | Accepted at 19th Conference on Neurosymbolic Learning and Reasoning (NeSy 2025) | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) often struggle to perform multi-target reasoning in long-context scenarios where relevant information is scattered across extensive documents. To address this challenge, we introduce NeuroSymbolic Augmented Reasoning (NSAR), which combines the benefits of neural and symbolic reasoning during inference. NSAR explicitly extracts symbolic facts from text and generates executable Python code to handle complex reasoning steps. Through extensive experiments across seven languages and diverse context lengths, we demonstrate that NSAR significantly outperforms both a vanilla RAG baseline and advanced prompting strategies in accurately identifying and synthesizing multiple pieces of information. Our results highlight the effectiveness of combining explicit symbolic operations with neural inference for robust, interpretable, and scalable reasoning in multilingual settings. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:54:20 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Nezhad",
"Sina Bagheri",
""
],
[
"Agrawal",
"Ameeta",
""
]
] |
2506.02484 | Yasuhiro Shimizu | Kanako Doi, Hayase Takei, Yusaku Shinoda, Yoshihiko Okamoto, Daigorou Hirai, Koshi Takenaka, Taku Matsushita, Yoshiaki Kobayashi, Yasuhiro Shimizu | Anisotropic superconducting gap probed by $^{125}$Te NMR in noncentrosymmetric Sc$_6M$Te$_2$ ($M$ = Fe, Co) | 7 pages, 7 figures | Phys. Rev. B. 111, 214502 (2025) | 10.1103/PhysRevB.111.214502 | null | cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | The superconducting gap symmetry is investigated by $^{125}$Te NMR measurements on Sc$_6M$Te$_2$ ($M$ = Fe, Co) without spatial inversion symmetry. The spin susceptibility obtained from the Knight shift $K$ is suppressed below the superconducting transition temperature, while leaving a finite value down to the lowest temperature ($\simeq 0.4$ K). The nuclear spin-lattice relaxation rate $1/T_1$ follows a power law against temperature $T$ without showing a coherence peak characteristic of the isotropic gap. The result implies a pairing admixture or a residual density of states under magnetic field. The normal metallic state has a Korringa scaling relation between $1/T_1T$ and the Knight shift, reflecting a weak electron correlation. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:54:28 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Doi",
"Kanako",
""
],
[
"Takei",
"Hayase",
""
],
[
"Shinoda",
"Yusaku",
""
],
[
"Okamoto",
"Yoshihiko",
""
],
[
"Hirai",
"Daigorou",
""
],
[
"Takenaka",
"Koshi",
""
],
[
"Matsushita",
"Taku",
""
],
[
"Kobayashi",
"Yoshiaki",
""
],
[
"Shimizu",
"Yasuhiro",
""
]
] |
2506.02485 | Haowen Xu | Haowen Xu, Sisi Zlatanova, Ruiyu Liang, Ismet Canbulat | Generative AI for Predicting 2D and 3D Wildfire Spread: Beyond Physics-Based Models and Traditional Deep Learning | null | null | null | null | cs.AI cs.CE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Wildfires continue to inflict devastating human, environmental, and economic losses globally, as tragically exemplified by the 2025 Los Angeles wildfire and the urgent demand for more effective response strategies. While physics-based and deep learning models have advanced wildfire simulation, they face critical limitations in predicting and visualizing multimodal fire spread in real time, particularly in both 2D and 3D spatial domains using dynamically updated GIS data. These limitations hinder timely emergency response, infrastructure protection, and community safety. Generative AI has recently emerged as a transformative approach across research and industry. Models such as Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), Transformers, and diffusion-based architectures offer distinct advantages over traditional methods, including the integration of multimodal data, generation of diverse scenarios under uncertainty, and improved modeling of wildfire dynamics across spatial and temporal scales. This position paper advocates for the adoption of generative AI as a foundational framework for wildfire prediction. We explore how such models can enhance 2D fire spread forecasting and enable more realistic, scalable 3D simulations. Additionally, we employ a novel human-AI collaboration framework using large language models (LLMs) for automated knowledge extraction, literature synthesis, and bibliometric mapping. Looking ahead, we identify five key visions for integrating generative AI into wildfire management: multimodal approaches, AI foundation models, conversational AI systems, edge-computing-based scenario generation, and cognitive digital twins. We also address three major challenges accompanying these opportunities and propose potential solutions to support their implementation. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:54:40 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Xu",
"Haowen",
""
],
[
"Zlatanova",
"Sisi",
""
],
[
"Liang",
"Ruiyu",
""
],
[
"Canbulat",
"Ismet",
""
]
] |
2506.02486 | Baodi Shan | Baodi Shan, Mauricio Arayr-Polo, Barbara Chapman | DiOMP-Offloading: Toward Portable Distributed Heterogeneous OpenMP | null | null | null | null | cs.DC | http://creativecommons.org/licenses/by/4.0/ | As core counts and heterogeneity rise in HPC, traditional hybrid programming models face challenges in managing distributed GPU memory and ensuring portability. This paper presents DiOMP, a distributed OpenMP framework that unifies OpenMP target offloading with the Partitioned Global Address Space (PGAS) model. Built atop LLVM/OpenMP and using GASNet-EX or GPI-2 for communication, DiOMP transparently handles global memory, supporting both symmetric and asymmetric GPU allocations. It leverages OMPCCL, a portable collective communication layer compatible with vendor libraries. DiOMP simplifies programming by abstracting device memory and communication, achieving superior scalability and programmability over traditional approaches. Evaluations on NVIDIA A100, Grace Hopper, and AMD MI250X show improved performance in micro-benchmarks and applications like matrix multiplication and Minimod, highlighting DiOMP's potential for scalable, portable, and efficient heterogeneous computing. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:57:06 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Shan",
"Baodi",
""
],
[
"Arayr-Polo",
"Mauricio",
""
],
[
"Chapman",
"Barbara",
""
]
] |
2506.02487 | Jacob Kurlander | Jacob A. Kurlander, Pedro H. Bernardinelli, Megan E. Schwamb, Mario Juric, Joseph Murtagh, Colin Orion Chandler, Stephanie R. Merritt, David Nesvorny, David Vokrouhlicky, R. Lynne Jones, Grigori Fedorets, Samuel Cornwall, Matthew J. Holman, Siegfried Eggl, Drew Oldag, Maxine West, Jeremy Kubica, Peter Yoachim, Joachim Moeyens, Kathleen Kiker, Laura E. Buchanan | Predictions of the LSST Solar System Yield: Near-Earth Objects, Main Belt Asteroids, Jupiter Trojans, and Trans-Neptunian Objects | In press at The Astronomical Journal | null | 10.3847/1538-3881/add685 | null | astro-ph.EP astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The NSF-DOE Vera C. Rubin Observatory is a new 8m-class survey facility presently being commissioned in Chile, expected to begin the 10yr-long Legacy Survey of Space and Time (LSST) by the end of 2025. Using the purpose-built Sorcha survey simulator (Merritt et al. In Press), and near-final observing cadence, we perform the first high-fidelity simulation of LSST's solar system catalog for key small body populations. We show that the final LSST catalog will deliver over 1.1 billion observations of small bodies and raise the number of known objects to 1.27E5 near-Earth objects, 5.09E6 main belt asteroids, 1.09E5 Jupiter Trojans, and 3.70E4 trans-Neptunian objects. These represent 4-9x more objects than are presently known in each class, making LSST the largest source of data for small body science in this and the following decade. We characterize the measurements available for these populations, including orbits, griz colors, and lightcurves, and point out science opportunities they open. Importantly, we show that ~70% of the main asteroid belt and more distant populations will be discovered in the first two years of the survey, making high-impact solar system science possible from very early on. We make our simulated LSST catalog publicly available, allowing researchers to test their methods on an up-to-date, representative, full-scale simulation of LSST data. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 05:58:20 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kurlander",
"Jacob A.",
""
],
[
"Bernardinelli",
"Pedro H.",
""
],
[
"Schwamb",
"Megan E.",
""
],
[
"Juric",
"Mario",
""
],
[
"Murtagh",
"Joseph",
""
],
[
"Chandler",
"Colin Orion",
""
],
[
"Merritt",
"Stephanie R.",
""
],
[
"Nesvorny",
"David",
""
],
[
"Vokrouhlicky",
"David",
""
],
[
"Jones",
"R. Lynne",
""
],
[
"Fedorets",
"Grigori",
""
],
[
"Cornwall",
"Samuel",
""
],
[
"Holman",
"Matthew J.",
""
],
[
"Eggl",
"Siegfried",
""
],
[
"Oldag",
"Drew",
""
],
[
"West",
"Maxine",
""
],
[
"Kubica",
"Jeremy",
""
],
[
"Yoachim",
"Peter",
""
],
[
"Moeyens",
"Joachim",
""
],
[
"Kiker",
"Kathleen",
""
],
[
"Buchanan",
"Laura E.",
""
]
] |
2506.02489 | Tao Zhong | Tao Zhong, Jonah Buchanan, Christine Allen-Blanchette | Grasp2Grasp: Vision-Based Dexterous Grasp Translation via Schr\"odinger Bridges | 19 pages, 4 figures | null | null | null | cs.RO cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | We propose a new approach to vision-based dexterous grasp translation, which aims to transfer grasp intent across robotic hands with differing morphologies. Given a visual observation of a source hand grasping an object, our goal is to synthesize a functionally equivalent grasp for a target hand without requiring paired demonstrations or hand-specific simulations. We frame this problem as a stochastic transport between grasp distributions using the Schr\"odinger Bridge formalism. Our method learns to map between source and target latent grasp spaces via score and flow matching, conditioned on visual observations. To guide this translation, we introduce physics-informed cost functions that encode alignment in base pose, contact maps, wrench space, and manipulability. Experiments across diverse hand-object pairs demonstrate our approach generates stable, physically grounded grasps with strong generalization. This work enables semantic grasp transfer for heterogeneous manipulators and bridges vision-based grasping with probabilistic generative modeling. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:08:51 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zhong",
"Tao",
""
],
[
"Buchanan",
"Jonah",
""
],
[
"Allen-Blanchette",
"Christine",
""
]
] |
2506.02490 | Yong Xiang | Yong Xiang (1), Charley Peter Chen (2), Liyi Zeng (3), Wei Yin (1), Xin Liu (1), Hu Li (4), Wei Xu (1) ((1) Tsinghua University, (2) Harmonic Inc, (3) Peng Cheng Laboratory, (4) Unaffiliated) | Simplifying Root Cause Analysis in Kubernetes with StateGraph and LLM | 12 pages, 13 figures, 5 tables | null | null | null | cs.DC cs.AI cs.SE | http://creativecommons.org/licenses/by/4.0/ | Kubernetes, a notably complex and distributed system, utilizes an array of controllers to uphold cluster management logic through state reconciliation. Nevertheless, maintaining state consistency presents significant challenges due to unexpected failures, network disruptions, and asynchronous issues, especially within dynamic cloud environments. These challenges result in operational disruptions and economic losses, underscoring the necessity for robust root cause analysis (RCA) to enhance Kubernetes reliability. The development of large language models (LLMs) presents a promising direction for RCA. However, existing methodologies encounter several obstacles, including the diverse and evolving nature of Kubernetes incidents, the intricate context of incidents, and the polymorphic nature of these incidents. In this paper, we introduce SynergyRCA, an innovative tool that leverages LLMs with retrieval augmentation from graph databases and enhancement with expert prompts. SynergyRCA constructs a StateGraph to capture spatial and temporal relationships and utilizes a MetaGraph to outline entity connections. Upon the occurrence of an incident, an LLM predicts the most pertinent resource, and SynergyRCA queries the MetaGraph and StateGraph to deliver context-specific insights for RCA. We evaluate SynergyRCA using datasets from two production Kubernetes clusters, highlighting its capacity to identify numerous root causes, including novel ones, with high efficiency and precision. SynergyRCA demonstrates the ability to identify root causes in an average time of about two minutes and achieves an impressive precision of approximately 0.90. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:09:13 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Xiang",
"Yong",
"",
"Tsinghua University"
],
[
"Chen",
"Charley Peter",
"",
"Harmonic Inc"
],
[
"Zeng",
"Liyi",
"",
"Peng Cheng Laboratory"
],
[
"Yin",
"Wei",
"",
"Tsinghua University"
],
[
"Liu",
"Xin",
"",
"Tsinghua University"
],
[
"Li",
"Hu",
"",
"Unaffiliated"
],
[
"Xu",
"Wei",
"",
"Tsinghua University"
]
] |
2506.02492 | Yuanpeng He | Yuanpeng He, Lijian Li, Tianxiang Zhan, Chi-Man Pun, Wenpin Jiao, Zhi Jin | Co-Evidential Fusion with Information Volume for Medical Image Segmentation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although existing semi-supervised image segmentation methods have achieved good performance, they cannot effectively utilize multiple sources of voxel-level uncertainty for targeted learning. Therefore, we propose two main improvements. First, we introduce a novel pignistic co-evidential fusion strategy using generalized evidential deep learning, extended by traditional D-S evidence theory, to obtain a more precise uncertainty measure for each voxel in medical samples. This assists the model in learning mixed labeled information and establishing semantic associations between labeled and unlabeled data. Second, we introduce the concept of information volume of mass function (IVUM) to evaluate the constructed evidence, implementing two evidential learning schemes. One optimizes evidential deep learning by combining the information volume of the mass function with original uncertainty measures. The other integrates the learning pattern based on the co-evidential fusion strategy, using IVUM to design a new optimization objective. Experiments on four datasets demonstrate the competitive performance of our method. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:13:19 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"He",
"Yuanpeng",
""
],
[
"Li",
"Lijian",
""
],
[
"Zhan",
"Tianxiang",
""
],
[
"Pun",
"Chi-Man",
""
],
[
"Jiao",
"Wenpin",
""
],
[
"Jin",
"Zhi",
""
]
] |
2506.02493 | Rui Yu | Jiachen Liu, Rui Yu, Sili Chen, Sharon X. Huang, Hengkai Guo | Towards In-the-wild 3D Plane Reconstruction from a Single Image | CVPR 2025 Highlighted Paper | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | 3D plane reconstruction from a single image is a crucial yet challenging topic in 3D computer vision. Previous state-of-the-art (SOTA) methods have focused on training their system on a single dataset from either indoor or outdoor domain, limiting their generalizability across diverse testing data. In this work, we introduce a novel framework dubbed ZeroPlane, a Transformer-based model targeting zero-shot 3D plane detection and reconstruction from a single image, over diverse domains and environments. To enable data-driven models across multiple domains, we have curated a large-scale planar benchmark, comprising over 14 datasets and 560,000 high-resolution, dense planar annotations for diverse indoor and outdoor scenes. To address the challenge of achieving desirable planar geometry on multi-dataset training, we propose to disentangle the representation of plane normal and offset, and employ an exemplar-guided, classification-then-regression paradigm to learn plane and offset respectively. Additionally, we employ advanced backbones as image encoder, and present an effective pixel-geometry-enhanced plane embedding module to further facilitate planar reconstruction. Extensive experiments across multiple zero-shot evaluation datasets have demonstrated that our approach significantly outperforms previous methods on both reconstruction accuracy and generalizability, especially over in-the-wild data. Our code and data are available at: https://github.com/jcliu0428/ZeroPlane. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:14:05 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Liu",
"Jiachen",
""
],
[
"Yu",
"Rui",
""
],
[
"Chen",
"Sili",
""
],
[
"Huang",
"Sharon X.",
""
],
[
"Guo",
"Hengkai",
""
]
] |
2506.02494 | Junzhe Zhang | Junzhe Zhang, Huixuan Zhang, Xinyu Hu, Li Lin, Mingqi Gao, Shi Qiu, Xiaojun Wan | Minos: A Multimodal Evaluation Model for Bidirectional Generation Between Image and Text | null | null | null | null | cs.CL cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Evaluation is important for multimodal generation tasks. With the rapid progress of MLLMs, there is growing interest in applying MLLMs to build general evaluation systems. However, existing work overlooks two aspects: (1) the development of evaluation capabilities for text-to-image (T2I) generation task, and (2) the incorporation of large-scale human evaluation data. In this paper, we introduce Minos-Corpus, a large-scale multimodal evaluation dataset that combines evaluation data from both human and GPT. The corpus contains evaluation data across both image-to-text(I2T) and T2I generation tasks. Based on this corpus, we propose Data Selection and Balance, Mix-SFT training methods, and apply DPO to develop Minos, a multimodal evaluation model built upon a 7B backbone. Minos achieves state-of-the-art (SoTA) performance among all open-source evaluation models of similar scale on the average of evaluation performance on all tasks, and outperforms all open-source and closed-source models on evaluation of T2I generation task. Extensive experiments demonstrate the importance of leveraging high-quality human evaluation data and jointly training on evaluation data from both I2T and T2I generation tasks. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:17:16 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zhang",
"Junzhe",
""
],
[
"Zhang",
"Huixuan",
""
],
[
"Hu",
"Xinyu",
""
],
[
"Lin",
"Li",
""
],
[
"Gao",
"Mingqi",
""
],
[
"Qiu",
"Shi",
""
],
[
"Wan",
"Xiaojun",
""
]
] |
2506.02495 | I\~nigo Liberal | I\~nigo Liberal and Alejandro Manjavacas | Synthetic crystal rotation with spacetime metamaterials | null | null | null | null | physics.optics cond-mat.mes-hall quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The interaction of light with rotating bodies has been historically limited to rotation frequencies much smaller than optical frequencies. Here, we investigate synthetic crystal rotations, i.e., spatiotemporal modulations mimicking the rotation of an anisotropic crystal, which grant access to large rotation frequencies. Spatiotemporal modulations change the fundamental symmetries of the electromagnetic field, breaking temporal and rotation symmetries, but preserving a spatiotemporal rotation symmetry that enforces the conservation of a combination of energy and spin angular momentum (SAM). The scattering of optical pulses by synthetically rotating crystals results in spatiotemporal light with intra-pulse SAM changes. The frequency-domain response reveals sidebands with frequency/SAM locking, and negative frequency sideband transitions for large enough rotation frequencies. Our results highlight the qualitatively different light-matter interaction regimes accessed by synthetic rotations. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:20:27 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Liberal",
"Iñigo",
""
],
[
"Manjavacas",
"Alejandro",
""
]
] |
2506.02496 | Yuan Yao | Hang Su, Yuan Yao, Akira Furusaki | Identification of gapless phases by squaring a twist operator | 6 pages, 3 figures | null | null | null | cond-mat.str-el cond-mat.stat-mech hep-th math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a general necessary condition for a spin chain with SO(3) spin-rotation symmetry to be gapped. Specifically, we prove that the ground state(s) of an SO(3)-symmetric gapped spin chain must be spin singlet(s), and the expectation value of the square of a twist operator asymptotically approaches unity in the thermodynamic limit, where finite-size corrections are inversely proportional to the square root of the system size. This theorem provides (i) supporting evidence for various conjectured gapped phases, and (ii) a sufficient criterion for identifying gapless spin chains. We verify our theorem by numerical simulations for a variety of spin models and show that it offers a novel efficient way to identify gapless phases in spin chains with spin-rotation symmetry. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:23:39 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Su",
"Hang",
""
],
[
"Yao",
"Yuan",
""
],
[
"Furusaki",
"Akira",
""
]
] |
2506.02497 | Jiahao Chen | Jiahao Chen, Hangjie Yuan, Yichen Qian, Jingyun Liang, Jiazheng Xing, Pengwei Liu, Weihua Chen, Fan Wang, Bing Su | LumosFlow: Motion-Guided Long Video Generation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Long video generation has gained increasing attention due to its widespread applications in fields such as entertainment and simulation. Despite advances, synthesizing temporally coherent and visually compelling long sequences remains a formidable challenge. Conventional approaches often synthesize long videos by sequentially generating and concatenating short clips, or generating key frames and then interpolate the intermediate frames in a hierarchical manner. However, both of them still remain significant challenges, leading to issues such as temporal repetition or unnatural transitions. In this paper, we revisit the hierarchical long video generation pipeline and introduce LumosFlow, a framework introduce motion guidance explicitly. Specifically, we first employ the Large Motion Text-to-Video Diffusion Model (LMTV-DM) to generate key frames with larger motion intervals, thereby ensuring content diversity in the generated long videos. Given the complexity of interpolating contextual transitions between key frames, we further decompose the intermediate frame interpolation into motion generation and post-hoc refinement. For each pair of key frames, the Latent Optical Flow Diffusion Model (LOF-DM) synthesizes complex and large-motion optical flows, while MotionControlNet subsequently refines the warped results to enhance quality and guide intermediate frame generation. Compared with traditional video frame interpolation, we achieve 15x interpolation, ensuring reasonable and continuous motion between adjacent frames. Experiments show that our method can generate long videos with consistent motion and appearance. Code and models will be made publicly available upon acceptance. Our project page: https://jiahaochen1.github.io/LumosFlow/ | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:25:00 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Chen",
"Jiahao",
""
],
[
"Yuan",
"Hangjie",
""
],
[
"Qian",
"Yichen",
""
],
[
"Liang",
"Jingyun",
""
],
[
"Xing",
"Jiazheng",
""
],
[
"Liu",
"Pengwei",
""
],
[
"Chen",
"Weihua",
""
],
[
"Wang",
"Fan",
""
],
[
"Su",
"Bing",
""
]
] |
2506.02498 | Yuan Ji Xu Dr. | Yuanji Xu, Xintao Jin, Haoyuan Tang, Fuyang Tian | Electronic structures and magnetism in van der Waals flat-band material Ni$_{3}$GeTe$_{2}$ | null | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | The study of magnetism in two-dimensional materials has garnered significant interest, driven by fundamental investigations into low-dimensional magnetic phenomena and their potential for applications in spintronic devices. Through dynamical mean-field theory calculations, we demonstrate that Ni$_{3}$GeTe$_{2}$ exhibits flat-band characteristics resulting from the geometric frustration of its layered triangular lattice. These flat bands are further renormalized due to moderate electronic correlation. Our calculations reveal that the magnetic order of Ni atoms is significantly influenced by both the Coulomb interaction and Hund's coupling, indicating that the physics of Ni atoms is situated in an intermediate region between Hundness and Mottness. Additionally, our results show that Ni atoms experience significant spin fluctuations in their local moments, maintaining paramagnetism at low temperatures. Furthermore, we investigate the effect of vacancies, finding a substantial suppression of the density of states at the Fermi level. The physical mechanisms uncovered by our study provide a comprehensive understanding of the novel properties exhibited by this type of material. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:25:52 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Xu",
"Yuanji",
""
],
[
"Jin",
"Xintao",
""
],
[
"Tang",
"Haoyuan",
""
],
[
"Tian",
"Fuyang",
""
]
] |
2506.02499 | Takuya Hasumi | Takuya Hasumi, Yusuke Fujita | DnR-nonverbal: Cinematic Audio Source Separation Dataset Containing Non-Verbal Sounds | Accepted to Interspeech 2025, 5 pages, 3 figures, dataset is available at https://zenodo.org/records/15470640 | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a new dataset for cinematic audio source separation (CASS) that handles non-verbal sounds. Existing CASS datasets only contain reading-style sounds as a speech stem. These datasets differ from actual movie audio, which is more likely to include acted-out voices. Consequently, models trained on conventional datasets tend to have issues where emotionally heightened voices, such as laughter and screams, are more easily separated as an effect, not speech. To address this problem, we build a new dataset, DnR-nonverbal. The proposed dataset includes non-verbal sounds like laughter and screams in the speech stem. From the experiments, we reveal the issue of non-verbal sound extraction by the current CASS model and show that our dataset can effectively address the issue in the synthetic and actual movie audio. Our dataset is available at https://zenodo.org/records/15470640. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:25:53 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Hasumi",
"Takuya",
""
],
[
"Fujita",
"Yusuke",
""
]
] |
2506.02500 | Jianfu Zhang | Maichang Lei, Yuan Zheng, Jianfu Zhang, Yuhai Yuan, Jiancheng Wang | Insight into the origin of multiwavelength emissions of PKS 1510-089 through modeling 12 SEDs from 2008 to 2015 | 33 pages, 13 figure, 5 Tables. Accepted for publication in ApJ | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | PKS\,1510$-$089 is one of the most peculiar sources among the FSRQs, exhibiting a notable big blue bump (BBB). This provides an unique opportunity to explore the coupling between the activity of the central engine and the relativistic jet, offering further insight into the origin of the multiwavelength emissions. To this end, we collected multiwavelength data spanning four periods from 2008 to 2015 and performed the spectral energy distribution (SED) modeling using a one-zone homogeneous leptonic model. In the model, a multichromatic accretion disk (AD) is used to fit the optical/UV data sets, while the external radiation fields from the broad-line region (BLR) and dusty torus (DT) are properly considered to produce the high-energy $\gamma$-ray emissions. Our best fit to 12 SEDs yields the following results: (i) The innermost stable orbit ($R_{\rm ISO}$) of the AD is not stable but varies between $3\,R_{\rm S}$ and $18\,R_{\rm S}$ during these observations. (ii) The high-energy hump of the SED is well dominated by Compton scattering of the BLR photons, while the X-ray flux may be comprised of multiple radiation components. (iii) The $\gamma$-ray emitting regions are generally matter-dominated, with low magnetization, and are located beyond the BLR but within the DT. At such distance, the multiwavelength emissions are likely to originate from shock accelerations; (iv) For the energization of the relativistic jet, our study supports the Blandford$-$Znajek (BZ) mechanism, instead of the Blandford$-$Payne (BP) mechanism, as the latter fails to power the jet. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:29:12 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Lei",
"Maichang",
""
],
[
"Zheng",
"Yuan",
""
],
[
"Zhang",
"Jianfu",
""
],
[
"Yuan",
"Yuhai",
""
],
[
"Wang",
"Jiancheng",
""
]
] |
2506.02501 | Wance Wang | Wance Wang, Dhruv Fomra, Amit Agrawal, Henri J. Lezec, Joseph W. Britton | Can TCOs Transform Cavity-QED? | 12 pages, 7 figures | null | null | null | quant-ph physics.optics | http://creativecommons.org/licenses/by/4.0/ | Transparent conductive oxides (TCO) enable confinement of charge-sensitive ions and Rydberg atoms proximal to dielectric structures including waveguides and photon detectors. However, optical loss precludes the use of TCOs within high-finesse optical micro-resonators. Here we characterize a ZnO-based TCO that markedly reduces optical absorption. At 1650\text{ nm} we observe a 22,000 finesse in a Fabry-P\'erot optical cavity coated with a 30\text{ nm} ZnO layer. This is a 5000 times reduction relative to indium tin oxide (ITO) at this wavelength. The same ZnO film exhibits 0.01\text{ \ensuremath{\Omega}\ensuremath{\cdot}cm} surface resistivity at DC. We anticipate a step change in cavity-QED systems incorporating ultra-low loss TCOs like ZnO. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:29:32 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Wang",
"Wance",
""
],
[
"Fomra",
"Dhruv",
""
],
[
"Agrawal",
"Amit",
""
],
[
"Lezec",
"Henri J.",
""
],
[
"Britton",
"Joseph W.",
""
]
] |
2506.02502 | Susumu Kinpara | Susumu Kinpara | The Non-perturbative term for the Vector Form Factor of Pion Decay | 9 pages | null | null | null | hep-ph nucl-th | http://creativecommons.org/publicdomain/zero/1.0/ | The vector form factor of the decay $\pi^{+} \rightarrow \gamma + e^{+} + \nu_e$ is calculated by the method for the pseudovector pion-nucleon system. The non-perturbative term is taken into account by using the parameter for the self-energy of nucleon following our previous calculation of the pion form factor. The suppression of the anomalous interaction of proton in the loop integral is significant to understand the experimental value. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:30:49 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kinpara",
"Susumu",
""
]
] |
2506.02503 | Yongjian Li | Yongjian Li, HaoCheng Chu, Yukun Yan, Zhenghao Liu, Shi Yu, Zheni Zeng, Ruobing Wang, Sen Song, Zhiyuan Liu, Maosong Sun | KARE-RAG: Knowledge-Aware Refinement and Enhancement for RAG | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Retrieval-Augmented Generation (RAG) enables large language models (LLMs) to access broader knowledge sources, yet factual inconsistencies persist due to noise in retrieved documents-even with advanced retrieval methods. We demonstrate that enhancing generative models' capacity to process noisy content is equally critical for robust performance. In this paper, we present KARE-RAG (Knowledge-Aware Refinement and Enhancement for RAG), which improves knowledge utilization through three key innovations: (1) structured knowledge representations that facilitate error detection during training, (2) Dense Direct Preference Optimization (DDPO)-a refined training objective that prioritizes correction of critical errors, and (3) a contrastive data generation pipeline that maintains semantic consistency while rectifying factual inaccuracies. Experiments show our method significantly enhances standard RAG pipelines across model scales, improving both in-domain and out-of-domain task performance without compromising general capabilities. Notably, these gains are achieved with modest training data, suggesting data-efficient optimization is possible through targeted learning strategies. Our findings establish a new direction for RAG improvement: by improving how models learn to process retrieved content, we can enhance performance across diverse inference paradigms. All data and code will be publicly available on Github. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:31:17 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Li",
"Yongjian",
""
],
[
"Chu",
"HaoCheng",
""
],
[
"Yan",
"Yukun",
""
],
[
"Liu",
"Zhenghao",
""
],
[
"Yu",
"Shi",
""
],
[
"Zeng",
"Zheni",
""
],
[
"Wang",
"Ruobing",
""
],
[
"Song",
"Sen",
""
],
[
"Liu",
"Zhiyuan",
""
],
[
"Sun",
"Maosong",
""
]
] |
2506.02504 | Xingyu Chen | Xingyu Chen, Bokun Wang, Ming Yang, Quanqi Hu, Qihang Lin, Tianbao Yang | Stochastic Momentum Methods for Non-smooth Non-Convex Finite-Sum Coupled Compositional Optimization | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Finite-sum Coupled Compositional Optimization (FCCO), characterized by its coupled compositional objective structure, emerges as an important optimization paradigm for addressing a wide range of machine learning problems. In this paper, we focus on a challenging class of non-convex non-smooth FCCO, where the outer functions are non-smooth weakly convex or convex and the inner functions are smooth or weakly convex. Existing state-of-the-art result face two key limitations: (1) a high iteration complexity of $O(1/\epsilon^6)$ under the assumption that the stochastic inner functions are Lipschitz continuous in expectation; (2) reliance on vanilla SGD-type updates, which are not suitable for deep learning applications. Our main contributions are two fold: (i) We propose stochastic momentum methods tailored for non-smooth FCCO that come with provable convergence guarantees; (ii) We establish a new state-of-the-art iteration complexity of $O(1/\epsilon^5)$. Moreover, we apply our algorithms to multiple inequality constrained non-convex optimization problems involving smooth or weakly convex functional inequality constraints. By optimizing a smoothed hinge penalty based formulation, we achieve a new state-of-the-art complexity of $O(1/\epsilon^5)$ for finding an (nearly) $\epsilon$-level KKT solution. Experiments on three tasks demonstrate the effectiveness of the proposed algorithms. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:31:59 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Chen",
"Xingyu",
""
],
[
"Wang",
"Bokun",
""
],
[
"Yang",
"Ming",
""
],
[
"Hu",
"Quanqi",
""
],
[
"Lin",
"Qihang",
""
],
[
"Yang",
"Tianbao",
""
]
] |
2506.02505 | Gaoyang Dong | Gaoyang Dong, Zhicheng Zhang, Ping Sun, Minghui Zhang | Adaptive Differential Denoising for Respiratory Sounds Classification | accepted at Interspeech2025 | null | null | null | eess.AS cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automated respiratory sound classification faces practical challenges from background noise and insufficient denoising in existing systems.
We propose Adaptive Differential Denoising network, that integrates noise suppression and pathological feature preservation via three innovations:
1) Adaptive Frequency Filter with learnable spectral masks and soft shrink to eliminate noise while retaining diagnostic high-frequency components;
2) A Differential Denoise Layer using differential attention to reduce noise-induced variations through augmented sample comparisons;
3) A bias denoising loss jointly optimizing classification and robustness without clean labels.
Experiments on the ICBHI2017 dataset show that our method achieves 65.53\% of the Score, which is improved by 1.99\% over the previous sota method.
The code is available in https://github.com/deegy666/ADD-RSC | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:32:27 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Dong",
"Gaoyang",
""
],
[
"Zhang",
"Zhicheng",
""
],
[
"Sun",
"Ping",
""
],
[
"Zhang",
"Minghui",
""
]
] |
2506.02506 | Pralay Kumar Das | Pralay Kumar Das, Nayana Majumdar and Supratik Mukhopadhyay | Classification of Hoyle State Decay Branches in Active Target Time Projection Chamber using Neural Network | null | null | null | null | physics.ins-det nucl-ex physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A multi-class convolutional neural network (CNN) model has been developed using Keras deep learning library in Python for image-based classification of $^{12}$C Hoyle state decay branches from tracking information, recorded by Saha Active Target Time Projection Chamber, SAT-TPC (currently under development). The nuclear events, produced by the 30 MeV $\alpha$-particle beam in the SAT-TPC, filled with Ar + CO$_2$ (90:10) gas mixture at atmospheric pressure, have been considered for training and validation of the models. The elastic scattering and Hoyle state sequential and direct decay events in the interaction of $\alpha$-particle with $^{40}$Ar, $^{12}$C, $^{16}$O nuclei have been generated through Monte-Carlo simulation. The three-dimensional tracks, produced by the scattering and decay products through primary ionization of gaseous medium, have been simulated with Geant4. The primary tracks, distributed on the beam-plane, have been convoluted with electron diffusion, obtained with Magboltz, to produce the final tracking information. The classification performance of the proposed model for different readout segmentation schemes of the SAT-TPC has been discussed. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:35:24 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Das",
"Pralay Kumar",
""
],
[
"Majumdar",
"Nayana",
""
],
[
"Mukhopadhyay",
"Supratik",
""
]
] |
2506.02507 | Alvin Zhu | Alvin Zhu, Yusuke Tanaka, Dennis Hong | AURA: Agentic Upskilling via Reinforced Abstractions | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the combinatorial explosion involved in translating high-level task prompts into deployable control policies for agile robots through multi-stage reinforcement learning. We introduce AURA (Agentic Upskilling via Reinforced Abstractions), a schema-centric curriculum RL framework that leverages Large Language Models (LLMs) as autonomous designers of multi-stage curricula. AURA transforms user prompts into YAML workflows that encode full reward functions, domain randomization strategies, and training configurations. All files are statically validated against a schema before any GPU time is consumed, ensuring reliable and efficient execution without human intervention. A retrieval-augmented feedback loop allows specialized LLM agents to design, execute, and refine staged curricula based on prior training results stored in a vector database, supporting continual improvement over time. Ablation studies highlight the importance of retrieval for curriculum quality and convergence stability. Quantitative experiments show that AURA consistently outperforms LLM-guided baselines on GPU-accelerated training frameworks. In qualitative tests, AURA successfully trains end-to-end policies directly from user prompts and deploys them zero-shot on a custom humanoid robot across a range of environments. By abstracting away the complexity of curriculum design, AURA enables scalable and adaptive policy learning pipelines that would be prohibitively complex to construct by hand. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:37:34 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zhu",
"Alvin",
""
],
[
"Tanaka",
"Yusuke",
""
],
[
"Hong",
"Dennis",
""
]
] |
2506.02508 | Telem Ibungochouba Singh | Yenshembam Priyobarta Singh, Jayasri Choudhury, Telem Ibungochouba Singh, Dhruba Jyoti Gogoi | Quasinormal modes and greybody factors of charge black hole in bumblebee gravity model | null | null | null | null | hep-ph | http://creativecommons.org/licenses/by-sa/4.0/ | In this paper, we investigate the Dirac field, scalar field and electromagnetic field perturbations of Reissner-Nordstorm-de Sitter (RNdS) and Reissner-Nordstorm-(anti)-de Sitter (RNAdS)-like black holes within the frame work of Einstein-bumblebee gravity. The effective potential, greybody factor and quasinormal modes (QNMs) are also explored by using Dirac equation, Klein-Gordon equation and Maxwell's equation. We find that for RNdS-like black hole increasing the Lorentz violation parameter $L$ consistently leads to decrease in the effective potential for all types of perturbations but for RNAdS case the influence of $L$ varies depending on the types of perturbation. Further for both RNdS and RNAdS-like black holes, increasing charge $Q$ reduces the effective potential in all the perturbations. The greybody factors of all the types of perturbations are also discussed. The results show that the greybody factors depend on the shape of the effective potential: higher (lower) potentials gives lower (higher) greybody factors. The QNMs frequencies of RNdS-like black hole for the massless field perturbations are discussed by using 6th order WKB approximation and Pad\'e approximation. We also analyze the time-domain profiles of the perturbations. The effects of Lorentz violation parameter $L$ and charge $Q$ to the photon sphere radius and shadow radius are also discussed. It is noted that increasing $Q$ and $L$ reduce the rise of shadow radius for RNdS-like black hole. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:37:35 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Singh",
"Yenshembam Priyobarta",
""
],
[
"Choudhury",
"Jayasri",
""
],
[
"Singh",
"Telem Ibungochouba",
""
],
[
"Gogoi",
"Dhruba Jyoti",
""
]
] |
2506.02509 | Jiajie Fu | Jiajie Fu, Haitong Tang, Arijit Khan, Sharad Mehrotra, Xiangyu Ke and Yunjun Gao | In-context Clustering-based Entity Resolution with Large Language Models: A Design Space Exploration | Accept by SIGMOD26 | null | null | null | cs.DB | http://creativecommons.org/licenses/by/4.0/ | Entity Resolution (ER) is a fundamental data quality improvement task that identifies and links records referring to the same real-world entity. Traditional ER approaches often rely on pairwise comparisons, which can be costly in terms of time and monetary resources, especially with large datasets. Recently, Large Language Models (LLMs) have shown promising results in ER tasks. However, existing methods typically focus on pairwise matching, missing the potential of LLMs to perform clustering directly in a more cost-effective and scalable manner. In this paper, we propose a novel in-context clustering approach for ER, where LLMs are used to cluster records directly, reducing both time complexity and monetary costs. We systematically investigate the design space for in-context clustering, analyzing the impact of factors such as set size, diversity, variation, and ordering of records on clustering performance. Based on these insights, we develop LLM-CER (LLM-powered Clustering-based ER), which achieves high-quality ER results while minimizing LLM API calls. Our approach addresses key challenges, including efficient cluster merging and LLM hallucination, providing a scalable and effective solution for ER. Extensive experiments on nine real-world datasets demonstrate that our method significantly improves result quality, achieving up to 150% higher accuracy, 10% increase in the F-measure, and reducing API calls by up to 5 times, while maintaining comparable monetary cost to the most cost-effective baseline. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:39:18 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Fu",
"Jiajie",
""
],
[
"Tang",
"Haitong",
""
],
[
"Khan",
"Arijit",
""
],
[
"Mehrotra",
"Sharad",
""
],
[
"Ke",
"Xiangyu",
""
],
[
"Gao",
"Yunjun",
""
]
] |
2506.02510 | Jie Zhu | Jie Zhu, Junhui Li, Yalong Wen, Xiandong Li, Lifan Guo, Feng Chen | M$^3$FinMeeting: A Multilingual, Multi-Sector, and Multi-Task Financial Meeting Understanding Evaluation Dataset | Accepted by ACL-2025 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent breakthroughs in large language models (LLMs) have led to the development of new benchmarks for evaluating their performance in the financial domain. However, current financial benchmarks often rely on news articles, earnings reports, or announcements, making it challenging to capture the real-world dynamics of financial meetings. To address this gap, we propose a novel benchmark called $\texttt{M$^3$FinMeeting}$, which is a multilingual, multi-sector, and multi-task dataset designed for financial meeting understanding. First, $\texttt{M$^3$FinMeeting}$ supports English, Chinese, and Japanese, enhancing comprehension of financial discussions in diverse linguistic contexts. Second, it encompasses various industry sectors defined by the Global Industry Classification Standard (GICS), ensuring that the benchmark spans a broad range of financial activities. Finally, $\texttt{M$^3$FinMeeting}$ includes three tasks: summarization, question-answer (QA) pair extraction, and question answering, facilitating a more realistic and comprehensive evaluation of understanding. Experimental results with seven popular LLMs reveal that even the most advanced long-context models have significant room for improvement, demonstrating the effectiveness of $\texttt{M$^3$FinMeeting}$ as a benchmark for assessing LLMs' financial meeting comprehension skills. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:41:09 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zhu",
"Jie",
""
],
[
"Li",
"Junhui",
""
],
[
"Wen",
"Yalong",
""
],
[
"Li",
"Xiandong",
""
],
[
"Guo",
"Lifan",
""
],
[
"Chen",
"Feng",
""
]
] |
2506.02511 | Kurando Baba | Kurando Baba and Osamu Ikawa | Compact symmetric triads and symmetric triads with multiplicities | 44 pages | null | null | null | math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we develop the theory of symmetric triads with multiplicities. First, we classify abstract symmetric triads with multiplicities. Second, we determine the symmetric triads with multiplicities corresponding to commutative compact symmetric triads. As applications, we give the classifications for commutative compact symmetric triads, which consist of two types depending on the choice of the equivalence relations. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:43:13 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Baba",
"Kurando",
""
],
[
"Ikawa",
"Osamu",
""
]
] |
2506.02512 | Sven Wiesner | Torsten Hoge, Shota Maehara, Sven Wiesner | Extendability of the $B_2$-arrangement | 23 pages | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | Let $(\mathscr{A},m)$ be a free multiarrangement, and let $\mathscr{E}$ be an extension of $(\mathscr{A},m)$. It is well known that if $\mathscr{A}$ is the Coxeter arrangement of type $A_2$, then a free extension of $(\mathscr{A},m)$ always exists. In this work, we demonstrate that if $\mathscr{A}$ is the Coxeter arrangement of type $B_2$, there exist infinitely many multiplicities for which no free extension of $(\mathscr{A},m)$ exists. This result has immediate consequences for the existence of free extensions in higher rank. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:43:27 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Hoge",
"Torsten",
""
],
[
"Maehara",
"Shota",
""
],
[
"Wiesner",
"Sven",
""
]
] |
2506.02513 | Hiromichi Nakazato | Hiromichi Nakazato and Tohru Ozawa | Characterization of the D'Alembertian by the Poincar\'e Invariance | 15 pages | Proceedings of the conference "Critical Phenomena in Nonlinear Partial Differential Equations, Harmonic Analysis, and Functional Inequalities" 2025 | null | null | math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | Many physical models are described by partial differential equations and the most important mathematical structure of the equations is governed by the corresponding linear partial differential operators. Those linear partial differential operators are sometimes determined by the symmetry under the group of motion. In this paper, the d'Alembertian is shown to be characterized as the only linear partial differential operator of the second order that is invariant under the Poincar\'e group and dilations in the Minkowski space-time $\mathbb R\times\mathbb R^n$. The method of proof depends on the analysis of the invariance of the corresponding polynomial in space-time under the time reflections and space rotations. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:43:48 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Nakazato",
"Hiromichi",
""
],
[
"Ozawa",
"Tohru",
""
]
] |
2506.02514 | Kyra Wang | Kyra Wang, Boon-Kiat Quek, Jessica Goh, Dorien Herremans | To Embody or Not: The Effect Of Embodiment On User Perception Of LLM-based Conversational Agents | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Embodiment in conversational agents (CAs) refers to the physical or visual representation of these agents, which can significantly influence user perception and interaction. Limited work has been done examining the effect of embodiment on the perception of CAs utilizing modern large language models (LLMs) in non-hierarchical cooperative tasks, a common use case of CAs as more powerful models become widely available for general use. To bridge this research gap, we conducted a mixed-methods within-subjects study on how users perceive LLM-based CAs in cooperative tasks when embodied and non-embodied. The results show that the non-embodied agent received significantly better quantitative appraisals for competence than the embodied agent, and in qualitative feedback, many participants believed that the embodied CA was more sycophantic than the non-embodied CA. Building on prior work on users' perceptions of LLM sycophancy and anthropomorphic features, we theorize that the typically-positive impact of embodiment on perception of CA credibility can become detrimental in the presence of sycophancy. The implication of such a phenomenon is that, contrary to intuition and existing literature, embodiment is not a straightforward way to improve a CA's perceived credibility if there exists a tendency to sycophancy. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:44:26 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Wang",
"Kyra",
""
],
[
"Quek",
"Boon-Kiat",
""
],
[
"Goh",
"Jessica",
""
],
[
"Herremans",
"Dorien",
""
]
] |
2506.02515 | Zhuohan Xie | Zhuohan Xie, Dhruv Sahnan, Debopriyo Banerjee, Georgi Georgiev, Rushil Thareja, Hachem Madmoun, Jinyan Su, Aaryamonvikram Singh, Yuxia Wang, Rui Xing, Fajri Koto, Haonan Li, Ivan Koychev, Tanmoy Chakraborty, Salem Lahlou, Veselin Stoyanov, Preslav Nakov | FinChain: A Symbolic Benchmark for Verifiable Chain-of-Thought Financial Reasoning | 15 pages, 8 figures, 2 tables | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Multi-step symbolic reasoning is critical for advancing downstream performance on financial tasks. Yet, benchmarks for systematically evaluating this capability are lacking. Existing datasets like FinQA and ConvFinQA supervise only final numerical answers, without assessing intermediate reasoning steps. To address this, we introduce FinChain, the first symbolic benchmark designed for verifiable Chain-of- Thought (CoT) financial reasoning. Spanning 54 topics across 12 financial domains, Fin- Chain offers five parameterized templates per topic, each varying in reasoning complexity and domain expertise required. Each dataset instance includes an executable Python trace, enabling automatic generation of extensive training data and easy adaptation to other domains. We also introduce ChainEval, a new metric for automatic evaluation of both final answers and intermediate reasoning. Benchmarking 30 LLMs on our dataset, we find that even state-of-the-art models have considerable room for improvement in multi-step financial reasoning. All templates and evaluation metrics for FinChain are available at https: //github.com/mbzuai-nlp/finchain. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:44:42 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Xie",
"Zhuohan",
""
],
[
"Sahnan",
"Dhruv",
""
],
[
"Banerjee",
"Debopriyo",
""
],
[
"Georgiev",
"Georgi",
""
],
[
"Thareja",
"Rushil",
""
],
[
"Madmoun",
"Hachem",
""
],
[
"Su",
"Jinyan",
""
],
[
"Singh",
"Aaryamonvikram",
""
],
[
"Wang",
"Yuxia",
""
],
[
"Xing",
"Rui",
""
],
[
"Koto",
"Fajri",
""
],
[
"Li",
"Haonan",
""
],
[
"Koychev",
"Ivan",
""
],
[
"Chakraborty",
"Tanmoy",
""
],
[
"Lahlou",
"Salem",
""
],
[
"Stoyanov",
"Veselin",
""
],
[
"Nakov",
"Preslav",
""
]
] |
2506.02516 | Bhanjan Debnath | Bhanjan Debnath and Parag Katira | Sequential binding-unbinding based specific interactions highlight exchange dynamics and size distribution of condensates | null | null | null | null | cond-mat.soft physics.bio-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Understanding the mechanisms of formation of organelles and biomolecular condensates via liquid-liquid phase separation is a topic of great interest. It is reasonable to assume that the formation of protein-specific condensates inside a crowded, multicomponent cellular environment requires specific protein-protein interactions with lifetimes longer than non-specific protein-protein interactions. These longer-lived, specific interactions can arise through complex protein binding and unbinding pathways. In this work, we investigate single-step and sequential multi-step mechanisms for protein binding-unbinding interactions and model the lifetimes of such interactions in a heuristic manner. We show that while similar average interaction lifetimes can be achieved when binding-unbinding follows a single-step or a sequential multi-step process, the two processes result in two distinct lifetime distributions: exponential and truncated power-law, respectively. We then combine this model with Brownian dynamics simulations to elucidate the impact of these distinct lifetime distributions on condensate dynamics. Our model qualitatively captures important features of condensates, such as their fluidic nature, aging in protein interactions and its impact on the exchange dynamics, and the size distribution of condensates. In a nutshell, our model unravels how specificity in protein binding-unbinding interaction pathways, resulting in unique interaction lifetime distributions, is a key factor shaping the physical properties of biomolecular condensates. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:45:01 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Debnath",
"Bhanjan",
""
],
[
"Katira",
"Parag",
""
]
] |
2506.02517 | Raghavendra Tripathi | Patrick Oliveira Santos, Raghavendra Tripathi, Pierre Youssef | Khintchine inequalities, trace monoids and Tur\'an-type problems | 16 pages+references | null | null | null | math.OA math.CO math.FA math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove scalar and operator-valued Khintchine inequalities for mixtures of free and tensor-independent semicircle variables, interpolating between classical and free Khintchine-type inequalities. Specifically, we characterize the norm of sums of $G$-independent semicircle variables in terms of the spectral radius of the Cayley graph associated with the trace monoid determined by the graph $G$. Our approach relies on a precise correspondence between closed paths in trace monoids and the norms of such operator sums. This correspondence uncovers connections between non-commutative probability, combinatorial group theory, and extremal graph theory. In particular, we formulate Tur\'an-type extremal problems that govern maximal norm growth under classical commutation constraints, and identify the extremal configurations. We hope that the methods and connections developed here will be useful in the study of non-commutative structures constrained by combinatorial symmetries. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:45:15 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Santos",
"Patrick Oliveira",
""
],
[
"Tripathi",
"Raghavendra",
""
],
[
"Youssef",
"Pierre",
""
]
] |
2506.02518 | Phuc H. Nguyen | Phuc H. Nguyen, Stephanie M. Engel, Amy H. Herring | Prenatal phthalate exposures and adiposity outcomes trajectories: a multivariate Bayesian factor regression approach | null | null | null | null | stat.ME stat.AP | http://creativecommons.org/licenses/by/4.0/ | We aim to assess the longitudinal effects of prenatal exposure to phthalates on the risk of childhood obesity in children aged 4 to 7, with potential time-varying and sex-specific effects. Multiple body-composition-related outcomes, such as BMI z-score, fat mass percentage, and waist circumference, are available in the data. Existing chemical mixture analyses often look at these outcomes individually due to the limited availability of multivariate models for mixture exposures. We propose a multivariate Bayesian factor regression that handles multicollinearity in chemical exposures and borrows information across highly correlated outcomes to improve estimation efficiency. We demonstrate the proposed method's utility through simulation studies and an analysis of data from the Mount Sinai Children's Environmental Health Study. We find the associations between prenatal phthalate exposures and adiposity outcomes in male children to be negative at early ages but to become positive as the children get older. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:47:12 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Nguyen",
"Phuc H.",
""
],
[
"Engel",
"Stephanie M.",
""
],
[
"Herring",
"Amy H.",
""
]
] |
2506.02519 | Sohan Patnaik | Sohan Patnaik, Milan Aggarwal, Sumit Bhatia, Balaji Krishnamurthy | Learning Together to Perform Better: Teaching Small-Scale LLMs to Collaborate via Preferential Rationale Tuning | Accepted at ACL Main 2025 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | LLMssuch as GPT-4 have shown a remarkable ability to solve complex questions by generating step-by-step rationales. Prior works have utilized this capability to improve smaller and cheaper LMs (say, with 7B parameters). However, various practical constraints, such as copyright and legal issues, owing to lack of transparency in the pre-training data of large (often closed) models, prevent their use in commercial settings. Little focus has been given to improving the innate reasoning ability of smaller models without distilling information from larger LLMs. To address this, we propose COLLATE, a trainable framework that tunes a (small) LLM to generate those outputs from a pool of diverse rationales that selectively improves the downstream task. COLLATE enforces multiple instances of the same LLM to exhibit distinct behavior and employs them to generate rationales to obtain diverse outputs. The LLM is then tuned via preference optimization to choose the candidate rationale which maximizes the likelihood of ground-truth answer. COLLATE outperforms several trainable and prompting baselines on 5 datasets across 3 domains: maths problem solving, natural language inference, and commonsense reasoning. We show the eff icacy of COLLATE on LLMs from different model families across varying parameter scales (1B to 8B) and demonstrate the benefit of multiple rationale providers guided by the end task through ablations. Code is released here (https://github.com/Sohanpatnaik106/collate). | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:50:08 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Patnaik",
"Sohan",
""
],
[
"Aggarwal",
"Milan",
""
],
[
"Bhatia",
"Sumit",
""
],
[
"Krishnamurthy",
"Balaji",
""
]
] |
2506.02521 | Liang Yan | BESIII Collaboration: M. Ablikim, M. N. Achasov, P. Adlarson, X. C. Ai, R. Aliberti, A. Amoroso, Q. An, Y. Bai, O. Bakina, Y. Ban, H.-R. Bao, V. Batozskaya, K. Begzsuren, N. Berger, M. Berlowski, M. Bertani, D. Bettoni, F. Bianchi, E. Bianco, A. Bortone, I. Boyko, R. A. Briere, A. Brueggemann, H. Cai, M. H. Cai, X. Cai, A. Calcaterra, G. F. Cao, N. Cao, S. A. Cetin, X. Y. Chai, J. F. Chang, G. R. Che, Y. Z. Che, G. Chelkov, C. H. Chen, Chao Chen, G. Chen, H. S. Chen, H. Y. Chen, M. L. Chen, S. J. Chen, S. L. Chen, S. M. Chen, T. Chen, X. R. Chen, X. T. Chen, Y. B. Chen, Y. Q. Chen, Y. Q. Chen, Z. J. Chen, Z. K. Chen, S. K. Choi, X. Chu, G. Cibinetto, F. Cossio, J. J. Cui, H. L. Dai, J. P. Dai, A. Dbeyssi, R. E. de Boer, D. Dedovich, C. Q. Deng, Z. Y. Deng, A. Denig, I. Denysenko, M. Destefanis, F. De Mori, B. Ding, X. X. Ding, Y. Ding, Y. Ding, Y. X. Ding, J. Dong, L. Y. Dong, M. Y. Dong, X. Dong, M. C. Du, S. X. Du, S. X. Du, Y. Y. Duan, Z. H. Duan, P. Egorov, G. F. Fan, J. J. Fan, Y. H. Fan, J. Fang, J. Fang, S. S. Fang, W. X. Fang, Y. Q. Fang, R. Farinelli, L. Fava, F. Feldbauer, G. Felici, C. Q. Feng, J. H. Feng, Y. T. Feng, M. Fritsch, C. D. Fu, J. L. Fu, Y. W. Fu, H. Gao, X. B. Gao, Y. N. Gao, Y. N. Gao, Y. Y. Gao, Yang Gao, S. Garbolino, I. Garzia, P. T. Ge, Z. W. Ge, C. Geng, E. M. Gersabeck, A. Gilman, K. Goetzen, J. D. Gong, L. Gong, W. X. Gong, W. Gradl, S. Gramigna, M. Greco, M. H. Gu, Y. T. Gu, C. Y. Guan, A. Q. Guo, L. B. Guo, M. J. Guo, R. P. Guo, Y. P. Guo, A. Guskov, J. Gutierrez, K. L. Han, T. T. Han, F. Hanisch, K. D. Hao, X. Q. Hao, F. A. Harris, K. K. He, K. L. He, F. H. Heinsius, C. H. Heinz, Y. K. Heng, C. Herold, T. Holtmann, P. C. Hong, G. Y. Hou, X. T. Hou, Y. R. Hou, Z. L. Hou, H. M. Hu, J. F. Hu, Q. P. Hu, S. L. Hu, T. Hu, Y. Hu, Z. M. Hu, G. S. Huang, K. X. Huang, L. Q. Huang, P. Huang, X. T. Huang, Y. P. Huang, Y. S. Huang, T. Hussain, N. H\"usken, N. in der Wiesche, J. Jackson, S. Janchiv, Q. Ji, Q. P. Ji, W. Ji, X. B. Ji, X. L. Ji, Y. Y. Ji, Z. K. Jia, D. Jiang, H. B. Jiang, P. C. Jiang, S. J. Jiang, T. J. Jiang, X. S. Jiang, Y. Jiang, J. B. Jiao, J. K. Jiao, Z. Jiao, S. Jin, Y. Jin, M. Q. Jing, X. M. Jing, T. Johansson, S. Kabana, N. Kalantar-Nayestanaki, X. L. Kang, X. S. Kang, M. Kavatsyuk, B. C. Ke, V. Khachatryan, A. Khoukaz, R. Kiuchi, O. B. Kolcu, B. Kopf, M. Kuessner, X. Kui, N. Kumar, A. Kupsc, W. K\"uhn, Q. Lan, W. N. Lan, T. T. Lei, M. Lellmann, T. Lenz, C. Li, C. Li, C. H. Li, C. K. Li, Cheng Li, D. M. Li, F. Li, G. Li, H. B. Li, H. J. Li, H. N. Li, Hui Li, J. R. Li, J. S. Li, K. Li, K. L. Li, K. L. Li, L. J. Li, Lei Li, M. H. Li, M. R. Li, P. L. Li, P. R. Li, Q. M. Li, Q. X. Li, R. Li, T. Li, T. Y. Li, W. D. Li, W. G. Li, X. Li, X. H. Li, X. L. Li, X. Y. Li, X. Z. Li, Y. Li, Y. G. Li, Y. P. Li, Z. J. Li, Z. Y. Li, C. Liang, H. Liang, Y. F. Liang, Y. T. Liang, G. R. Liao, L. B. Liao, M. H. Liao, Y. P. Liao, J. Libby, A. Limphirat, C. C. Lin, C. X. Lin, D. X. Lin, L. Q. Lin, T. Lin, B. J. Liu, B. X. Liu, C. Liu, C. X. Liu, F. Liu, F. H. Liu, Feng Liu, G. M. Liu, H. Liu, H. B. Liu, H. H. Liu, H. M. Liu, Huihui Liu, J. B. Liu, J. J. Liu, K. Liu, K. Liu, K. Y. Liu, Ke Liu, L. Liu, L. C. Liu, Lu Liu, P. L. Liu, Q. Liu, S. B. Liu, T. Liu, W. K. Liu, W. M. Liu, W. T. Liu, X. Liu, X. Liu, X. Y. Liu, Y. Liu, Y. Liu, Y. Liu, Y. B. Liu, Z. A. Liu, Z. D. Liu, Z. Q. Liu, X. C. Lou, F. X. Lu, H. J. Lu, J. G. Lu, X. L. Lu, Y. Lu, Y. H. Lu, Y. P. Lu, Z. H. Lu, C. L. Luo, J. R. Luo, J. S. Luo, M. X. Luo, T. Luo, X. L. Luo, Z. Y. Lv, X. R. Lyu, Y. F. Lyu, Y. H. Lyu, F. C. Ma, H. Ma, H. L. Ma, J. L. Ma, L. L. Ma, L. R. Ma, Q. M. Ma, R. Q. Ma, R. Y. Ma, T. Ma, X. T. Ma, X. Y. Ma, Y. M. Ma, F. E. Maas, I. MacKay, M. Maggiora, S. Malde, Y. J. Mao, Z. P. Mao, S. Marcello, F. M. Melendi, Y. H. Meng, Z. X. Meng, J. G. Messchendorp, G. Mezzadri, H. Miao, T. J. Min, R. E. Mitchell, X. H. Mo, B. Moses, N. Yu. Muchnoi, J. Muskalla, Y. Nefedov, F. Nerling, L. S. Nie, I. B. Nikolaev, Z. Ning, S. Nisar, Q. L. Niu, W. D. Niu, S. L. Olsen, Q. Ouyang, S. Pacetti, X. Pan, Y. Pan, A. Pathak, Y. P. Pei, M. Pelizaeus, H. P. Peng, Y. Y. Peng, K. Peters, J. L. Ping, R. G. Ping, S. Plura, V. Prasad, F. Z. Qi, H. R. Qi, M. Qi, S. Qian, W. B. Qian, C. F. Qiao, J. H. Qiao, J. J. Qin, J. L. Qin, L. Q. Qin, L. Y. Qin, P. B. Qin, X. P. Qin, X. S. Qin, Z. H. Qin, J. F. Qiu, Z. H. Qu, C. F. Redmer, A. Rivetti, M. Rolo, G. Rong, S. S. Rong, F. Rosini, Ch. Rosner, M. Q. Ruan, N. Salone, A. Sarantsev, Y. Schelhaas, K. Schoenning, M. Scodeggio, K. Y. Shan, W. Shan, X. Y. Shan, Z. J. Shang, J. F. Shangguan, L. G. Shao, M. Shao, C. P. Shen, H. F. Shen, W. H. Shen, X. Y. Shen, B. A. Shi, H. Shi, J. L. Shi, J. Y. Shi, S. Y. Shi, X. Shi, H. L. Song, J. J. Song, T. Z. Song, W. M. Song, Y. X. Song, S. Sosio, S. Spataro, F. Stieler, S. S Su, Y. J. Su, G. B. Sun, G. X. Sun, H. Sun, H. K. Sun, J. F. Sun, K. Sun, L. Sun, S. S. Sun, T. Sun, Y. C. Sun, Y. H. Sun, Y. J. Sun, Y. Z. Sun, Z. Q. Sun, Z. T. Sun, C. J. Tang, G. Y. Tang, J. Tang, L. F. Tang, M. Tang, Y. A. Tang, L. Y. Tao, M. Tat, J. X. Teng, J. Y. Tian, W. H. Tian, Y. Tian, Z. F. Tian, I. Uman, B. Wang, B. Wang, Bo Wang, C. Wang, Cong Wang, D. Y. Wang, H. J. Wang, J. J. Wang, K. Wang, L. L. Wang, L. W. Wang, M. Wang, M. Wang, N. Y. Wang, S. Wang, T. Wang, T. J. Wang, W. Wang, W. Wang, W. P. Wang, X. Wang, X. F. Wang, X. J. Wang, X. L. Wang, X. N. Wang, Y. Wang, Y. D. Wang, Y. F. Wang, Y. H. Wang, Y. L. Wang, Y. N. Wang, Y. Q. Wang, Yaqian Wang, Yi Wang, Yuan Wang, Z. Wang, Z. L. Wang, Z. L. Wang, Z. Q. Wang, Z. Y. Wang, D. H. Wei, H. R. Wei, F. Weidner, S. P. Wen, Y. R. Wen, U. Wiedner, G. Wilkinson, M. Wolke, C. Wu, J. F. Wu, L. H. Wu, L. J. Wu, L. J. Wu, Lianjie Wu, S. G. Wu, S. M. Wu, X. Wu, X. H. Wu, Y. J. Wu, Z. Wu, L. Xia, X. M. Xian, B. H. Xiang, D. Xiao, G. Y. Xiao, H. Xiao, Y. L. Xiao, Z. J. Xiao, C. Xie, K. J. Xie, X. H. Xie, Y. Xie, Y. G. Xie, Y. H. Xie, Z. P. Xie, T. Y. Xing, C. F. Xu, C. J. Xu, G. F. Xu, H. Y. Xu, H. Y. Xu, M. Xu, Q. J. Xu, Q. N. Xu, T. D. Xu, W. L. Xu, X. P. Xu, Y. Xu, Y. Xu, Y. C. Xu, Z. S. Xu, H. Y. Yan, L. Yan, W. B. Yan, W. C. Yan, W. P. Yan, X. Q. Yan, H. J. Yang, H. L. Yang, H. X. Yang, J. H. Yang, R. J. Yang, T. Yang, Y. Yang, Y. F. Yang, Y. H. Yang, Y. Q. Yang, Y. X. Yang, Y. Z. Yang, M. Ye, M. H. Ye, Z. J. Ye, Junhao Yin, Z. Y. You, B. X. Yu, C. X. Yu, G. Yu, J. S. Yu, M. C. Yu, T. Yu, X. D. Yu, Y. C. Yu, C. Z. Yuan, H. Yuan, J. Yuan, J. Yuan, L. Yuan, S. C. Yuan, Y. Yuan, Z. Y. Yuan, C. X. Yue, Ying Yue, A. A. Zafar, S. H. Zeng, X. Zeng, Y. Zeng, Y. J. Zeng, Y. J. Zeng, X. Y. Zhai, Y. H. Zhan, A. Q. Zhang, B. L. Zhang, B. X. Zhang, D. H. Zhang, G. Y. Zhang, G. Y. Zhang, H. Zhang, H. Zhang, H. C. Zhang, H. H. Zhang, H. Q. Zhang, H. R. Zhang, H. Y. Zhang, J. Zhang, J. Zhang, J. J. Zhang, J. L. Zhang, J. Q. Zhang, J. S. Zhang, J. W. Zhang, J. X. Zhang, J. Y. Zhang, J. Z. Zhang, Jianyu Zhang, L. M. Zhang, Lei Zhang, N. Zhang, P. Zhang, Q. Zhang, Q. Y. Zhang, R. Y. Zhang, S. H. Zhang, Shulei Zhang, X. M. Zhang, X. Y Zhang, X. Y. Zhang, Y. Zhang, Y. Zhang, Y. T. Zhang, Y. H. Zhang, Y. M. Zhang, Z. D. Zhang, Z. H. Zhang, Z. L. Zhang, Z. L. Zhang, Z. X. Zhang, Z. Y. Zhang, Z. Y. Zhang, Z. Z. Zhang, Zh. Zh. Zhang, G. Zhao, J. Y. Zhao, J. Z. Zhao, L. Zhao, Lei Zhao, M. G. Zhao, N. Zhao, R. P. Zhao, S. J. Zhao, Y. B. Zhao, Y. L. Zhao, Y. X. Zhao, Z. G. Zhao, A. Zhemchugov, B. Zheng, B. M. Zheng, J. P. Zheng, W. J. Zheng, X. R. Zheng, Y. H. Zheng, B. Zhong, C. Zhong, H. Zhou, J. Q. Zhou, J. Y. Zhou, S. Zhou, X. Zhou, X. K. Zhou, X. R. Zhou, X. Y. Zhou, Y. Z. Zhou, Z. C. Zhou, A. N. Zhu, J. Zhu, K. Zhu, K. J. Zhu, K. S. Zhu, L. Zhu, L. X. Zhu, S. H. Zhu, T. J. Zhu, W. D. Zhu, W. D. Zhu, W. J. Zhu, W. Z. Zhu, Y. C. Zhu, Z. A. Zhu, X. Y. Zhuang, J. H. Zou, J. Zu | Improved Measurements of $D^+ \to \eta e^+\nu_e$ and $D^+ \to \eta \mu^+\nu_\mu$ | null | null | null | null | hep-ex | http://creativecommons.org/licenses/by/4.0/ | Using 20.3 fb$^{-1}$ of $e^+e^-$ collision data collected at the center-of-mass energy of 3.773 GeV with the BESIII detector, we measure the branching fractions of $D^+\to \eta e^+\nu_e$ and $D^+\to \eta\mu^+\nu_\mu$ to be $(9.75\pm0.29\pm0.28)\times10^{-4}$ and $(9.08\pm0.35\pm0.23)\times10^{-4}$, where the first and second uncertainties are statistical and systematic, respectively. From a simultaneous fit to their partial decay rates, we determine the product of the hadronic form factor $f^\eta_+(0)$ and the modulus of the $c\to d$ Cabibbo-Kobayashi-Maskawa matrix element $|V_{cd}|$ to be $f^\eta_+(0)|V_{cd}|=0.078\pm0.002\pm0.001$. Taking the $|V_{cd}|$ value from the Standard Model global fit as input, we obtain $f^\eta_+(0)=0.345\pm0.008\pm0.003$. The ratio between the measured branching fractions of $D^+\to\eta^+\mu^+\nu_{\mu}$ and $D^+\to\eta e^+\nu_e$, is determined to be $0.93\pm0.05_{\rm stat.}\pm0.02_{\rm syst.}$, indicating no violation of lepton flavor universality. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:51:03 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"BESIII Collaboration",
"",
""
],
[
"Ablikim",
"M.",
""
],
[
"Achasov",
"M. N.",
""
],
[
"Adlarson",
"P.",
""
],
[
"Ai",
"X. C.",
""
],
[
"Aliberti",
"R.",
""
],
[
"Amoroso",
"A.",
""
],
[
"An",
"Q.",
""
],
[
"Bai",
"Y.",
""
],
[
"Bakina",
"O.",
""
],
[
"Ban",
"Y.",
""
],
[
"Bao",
"H. -R.",
""
],
[
"Batozskaya",
"V.",
""
],
[
"Begzsuren",
"K.",
""
],
[
"Berger",
"N.",
""
],
[
"Berlowski",
"M.",
""
],
[
"Bertani",
"M.",
""
],
[
"Bettoni",
"D.",
""
],
[
"Bianchi",
"F.",
""
],
[
"Bianco",
"E.",
""
],
[
"Bortone",
"A.",
""
],
[
"Boyko",
"I.",
""
],
[
"Briere",
"R. A.",
""
],
[
"Brueggemann",
"A.",
""
],
[
"Cai",
"H.",
""
],
[
"Cai",
"M. H.",
""
],
[
"Cai",
"X.",
""
],
[
"Calcaterra",
"A.",
""
],
[
"Cao",
"G. F.",
""
],
[
"Cao",
"N.",
""
],
[
"Cetin",
"S. A.",
""
],
[
"Chai",
"X. Y.",
""
],
[
"Chang",
"J. F.",
""
],
[
"Che",
"G. R.",
""
],
[
"Che",
"Y. Z.",
""
],
[
"Chelkov",
"G.",
""
],
[
"Chen",
"C. H.",
""
],
[
"Chen",
"Chao",
""
],
[
"Chen",
"G.",
""
],
[
"Chen",
"H. S.",
""
],
[
"Chen",
"H. Y.",
""
],
[
"Chen",
"M. L.",
""
],
[
"Chen",
"S. J.",
""
],
[
"Chen",
"S. L.",
""
],
[
"Chen",
"S. M.",
""
],
[
"Chen",
"T.",
""
],
[
"Chen",
"X. R.",
""
],
[
"Chen",
"X. T.",
""
],
[
"Chen",
"Y. B.",
""
],
[
"Chen",
"Y. Q.",
""
],
[
"Chen",
"Y. Q.",
""
],
[
"Chen",
"Z. J.",
""
],
[
"Chen",
"Z. K.",
""
],
[
"Choi",
"S. K.",
""
],
[
"Chu",
"X.",
""
],
[
"Cibinetto",
"G.",
""
],
[
"Cossio",
"F.",
""
],
[
"Cui",
"J. J.",
""
],
[
"Dai",
"H. L.",
""
],
[
"Dai",
"J. P.",
""
],
[
"Dbeyssi",
"A.",
""
],
[
"de Boer",
"R. E.",
""
],
[
"Dedovich",
"D.",
""
],
[
"Deng",
"C. Q.",
""
],
[
"Deng",
"Z. Y.",
""
],
[
"Denig",
"A.",
""
],
[
"Denysenko",
"I.",
""
],
[
"Destefanis",
"M.",
""
],
[
"De Mori",
"F.",
""
],
[
"Ding",
"B.",
""
],
[
"Ding",
"X. X.",
""
],
[
"Ding",
"Y.",
""
],
[
"Ding",
"Y.",
""
],
[
"Ding",
"Y. X.",
""
],
[
"Dong",
"J.",
""
],
[
"Dong",
"L. Y.",
""
],
[
"Dong",
"M. Y.",
""
],
[
"Dong",
"X.",
""
],
[
"Du",
"M. C.",
""
],
[
"Du",
"S. X.",
""
],
[
"Du",
"S. X.",
""
],
[
"Duan",
"Y. Y.",
""
],
[
"Duan",
"Z. H.",
""
],
[
"Egorov",
"P.",
""
],
[
"Fan",
"G. F.",
""
],
[
"Fan",
"J. J.",
""
],
[
"Fan",
"Y. H.",
""
],
[
"Fang",
"J.",
""
],
[
"Fang",
"J.",
""
],
[
"Fang",
"S. S.",
""
],
[
"Fang",
"W. X.",
""
],
[
"Fang",
"Y. Q.",
""
],
[
"Farinelli",
"R.",
""
],
[
"Fava",
"L.",
""
],
[
"Feldbauer",
"F.",
""
],
[
"Felici",
"G.",
""
],
[
"Feng",
"C. Q.",
""
],
[
"Feng",
"J. H.",
""
],
[
"Feng",
"Y. T.",
""
],
[
"Fritsch",
"M.",
""
],
[
"Fu",
"C. D.",
""
],
[
"Fu",
"J. L.",
""
],
[
"Fu",
"Y. W.",
""
],
[
"Gao",
"H.",
""
],
[
"Gao",
"X. B.",
""
],
[
"Gao",
"Y. N.",
""
],
[
"Gao",
"Y. N.",
""
],
[
"Gao",
"Y. Y.",
""
],
[
"Gao",
"Yang",
""
],
[
"Garbolino",
"S.",
""
],
[
"Garzia",
"I.",
""
],
[
"Ge",
"P. T.",
""
],
[
"Ge",
"Z. W.",
""
],
[
"Geng",
"C.",
""
],
[
"Gersabeck",
"E. M.",
""
],
[
"Gilman",
"A.",
""
],
[
"Goetzen",
"K.",
""
],
[
"Gong",
"J. D.",
""
],
[
"Gong",
"L.",
""
],
[
"Gong",
"W. X.",
""
],
[
"Gradl",
"W.",
""
],
[
"Gramigna",
"S.",
""
],
[
"Greco",
"M.",
""
],
[
"Gu",
"M. H.",
""
],
[
"Gu",
"Y. T.",
""
],
[
"Guan",
"C. Y.",
""
],
[
"Guo",
"A. Q.",
""
],
[
"Guo",
"L. B.",
""
],
[
"Guo",
"M. J.",
""
],
[
"Guo",
"R. P.",
""
],
[
"Guo",
"Y. P.",
""
],
[
"Guskov",
"A.",
""
],
[
"Gutierrez",
"J.",
""
],
[
"Han",
"K. L.",
""
],
[
"Han",
"T. T.",
""
],
[
"Hanisch",
"F.",
""
],
[
"Hao",
"K. D.",
""
],
[
"Hao",
"X. Q.",
""
],
[
"Harris",
"F. A.",
""
],
[
"He",
"K. K.",
""
],
[
"He",
"K. L.",
""
],
[
"Heinsius",
"F. H.",
""
],
[
"Heinz",
"C. H.",
""
],
[
"Heng",
"Y. K.",
""
],
[
"Herold",
"C.",
""
],
[
"Holtmann",
"T.",
""
],
[
"Hong",
"P. C.",
""
],
[
"Hou",
"G. Y.",
""
],
[
"Hou",
"X. T.",
""
],
[
"Hou",
"Y. R.",
""
],
[
"Hou",
"Z. L.",
""
],
[
"Hu",
"H. M.",
""
],
[
"Hu",
"J. F.",
""
],
[
"Hu",
"Q. P.",
""
],
[
"Hu",
"S. L.",
""
],
[
"Hu",
"T.",
""
],
[
"Hu",
"Y.",
""
],
[
"Hu",
"Z. M.",
""
],
[
"Huang",
"G. S.",
""
],
[
"Huang",
"K. X.",
""
],
[
"Huang",
"L. Q.",
""
],
[
"Huang",
"P.",
""
],
[
"Huang",
"X. T.",
""
],
[
"Huang",
"Y. P.",
""
],
[
"Huang",
"Y. S.",
""
],
[
"Hussain",
"T.",
""
],
[
"Hüsken",
"N.",
""
],
[
"der Wiesche",
"N. in",
""
],
[
"Jackson",
"J.",
""
],
[
"Janchiv",
"S.",
""
],
[
"Ji",
"Q.",
""
],
[
"Ji",
"Q. P.",
""
],
[
"Ji",
"W.",
""
],
[
"Ji",
"X. B.",
""
],
[
"Ji",
"X. L.",
""
],
[
"Ji",
"Y. Y.",
""
],
[
"Jia",
"Z. K.",
""
],
[
"Jiang",
"D.",
""
],
[
"Jiang",
"H. B.",
""
],
[
"Jiang",
"P. C.",
""
],
[
"Jiang",
"S. J.",
""
],
[
"Jiang",
"T. J.",
""
],
[
"Jiang",
"X. S.",
""
],
[
"Jiang",
"Y.",
""
],
[
"Jiao",
"J. B.",
""
],
[
"Jiao",
"J. K.",
""
],
[
"Jiao",
"Z.",
""
],
[
"Jin",
"S.",
""
],
[
"Jin",
"Y.",
""
],
[
"Jing",
"M. Q.",
""
],
[
"Jing",
"X. M.",
""
],
[
"Johansson",
"T.",
""
],
[
"Kabana",
"S.",
""
],
[
"Kalantar-Nayestanaki",
"N.",
""
],
[
"Kang",
"X. L.",
""
],
[
"Kang",
"X. S.",
""
],
[
"Kavatsyuk",
"M.",
""
],
[
"Ke",
"B. C.",
""
],
[
"Khachatryan",
"V.",
""
],
[
"Khoukaz",
"A.",
""
],
[
"Kiuchi",
"R.",
""
],
[
"Kolcu",
"O. B.",
""
],
[
"Kopf",
"B.",
""
],
[
"Kuessner",
"M.",
""
],
[
"Kui",
"X.",
""
],
[
"Kumar",
"N.",
""
],
[
"Kupsc",
"A.",
""
],
[
"Kühn",
"W.",
""
],
[
"Lan",
"Q.",
""
],
[
"Lan",
"W. N.",
""
],
[
"Lei",
"T. T.",
""
],
[
"Lellmann",
"M.",
""
],
[
"Lenz",
"T.",
""
],
[
"Li",
"C.",
""
],
[
"Li",
"C.",
""
],
[
"Li",
"C. H.",
""
],
[
"Li",
"C. K.",
""
],
[
"Li",
"Cheng",
""
],
[
"Li",
"D. M.",
""
],
[
"Li",
"F.",
""
],
[
"Li",
"G.",
""
],
[
"Li",
"H. B.",
""
],
[
"Li",
"H. J.",
""
],
[
"Li",
"H. N.",
""
],
[
"Li",
"Hui",
""
],
[
"Li",
"J. R.",
""
],
[
"Li",
"J. S.",
""
],
[
"Li",
"K.",
""
],
[
"Li",
"K. L.",
""
],
[
"Li",
"K. L.",
""
],
[
"Li",
"L. J.",
""
],
[
"Li",
"Lei",
""
],
[
"Li",
"M. H.",
""
],
[
"Li",
"M. R.",
""
],
[
"Li",
"P. L.",
""
],
[
"Li",
"P. R.",
""
],
[
"Li",
"Q. M.",
""
],
[
"Li",
"Q. X.",
""
],
[
"Li",
"R.",
""
],
[
"Li",
"T.",
""
],
[
"Li",
"T. Y.",
""
],
[
"Li",
"W. D.",
""
],
[
"Li",
"W. G.",
""
],
[
"Li",
"X.",
""
],
[
"Li",
"X. H.",
""
],
[
"Li",
"X. L.",
""
],
[
"Li",
"X. Y.",
""
],
[
"Li",
"X. Z.",
""
],
[
"Li",
"Y.",
""
],
[
"Li",
"Y. G.",
""
],
[
"Li",
"Y. P.",
""
],
[
"Li",
"Z. J.",
""
],
[
"Li",
"Z. Y.",
""
],
[
"Liang",
"C.",
""
],
[
"Liang",
"H.",
""
],
[
"Liang",
"Y. F.",
""
],
[
"Liang",
"Y. T.",
""
],
[
"Liao",
"G. R.",
""
],
[
"Liao",
"L. B.",
""
],
[
"Liao",
"M. H.",
""
],
[
"Liao",
"Y. P.",
""
],
[
"Libby",
"J.",
""
],
[
"Limphirat",
"A.",
""
],
[
"Lin",
"C. C.",
""
],
[
"Lin",
"C. X.",
""
],
[
"Lin",
"D. X.",
""
],
[
"Lin",
"L. Q.",
""
],
[
"Lin",
"T.",
""
],
[
"Liu",
"B. J.",
""
],
[
"Liu",
"B. X.",
""
],
[
"Liu",
"C.",
""
],
[
"Liu",
"C. X.",
""
],
[
"Liu",
"F.",
""
],
[
"Liu",
"F. H.",
""
],
[
"Liu",
"Feng",
""
],
[
"Liu",
"G. M.",
""
],
[
"Liu",
"H.",
""
],
[
"Liu",
"H. B.",
""
],
[
"Liu",
"H. H.",
""
],
[
"Liu",
"H. M.",
""
],
[
"Liu",
"Huihui",
""
],
[
"Liu",
"J. B.",
""
],
[
"Liu",
"J. J.",
""
],
[
"Liu",
"K.",
""
],
[
"Liu",
"K.",
""
],
[
"Liu",
"K. Y.",
""
],
[
"Liu",
"Ke",
""
],
[
"Liu",
"L.",
""
],
[
"Liu",
"L. C.",
""
],
[
"Liu",
"Lu",
""
],
[
"Liu",
"P. L.",
""
],
[
"Liu",
"Q.",
""
],
[
"Liu",
"S. B.",
""
],
[
"Liu",
"T.",
""
],
[
"Liu",
"W. K.",
""
],
[
"Liu",
"W. M.",
""
],
[
"Liu",
"W. T.",
""
],
[
"Liu",
"X.",
""
],
[
"Liu",
"X.",
""
],
[
"Liu",
"X. Y.",
""
],
[
"Liu",
"Y.",
""
],
[
"Liu",
"Y.",
""
],
[
"Liu",
"Y.",
""
],
[
"Liu",
"Y. B.",
""
],
[
"Liu",
"Z. A.",
""
],
[
"Liu",
"Z. D.",
""
],
[
"Liu",
"Z. Q.",
""
],
[
"Lou",
"X. C.",
""
],
[
"Lu",
"F. X.",
""
],
[
"Lu",
"H. J.",
""
],
[
"Lu",
"J. G.",
""
],
[
"Lu",
"X. L.",
""
],
[
"Lu",
"Y.",
""
],
[
"Lu",
"Y. H.",
""
],
[
"Lu",
"Y. P.",
""
],
[
"Lu",
"Z. H.",
""
],
[
"Luo",
"C. L.",
""
],
[
"Luo",
"J. R.",
""
],
[
"Luo",
"J. S.",
""
],
[
"Luo",
"M. X.",
""
],
[
"Luo",
"T.",
""
],
[
"Luo",
"X. L.",
""
],
[
"Lv",
"Z. Y.",
""
],
[
"Lyu",
"X. R.",
""
],
[
"Lyu",
"Y. F.",
""
],
[
"Lyu",
"Y. H.",
""
],
[
"Ma",
"F. C.",
""
],
[
"Ma",
"H.",
""
],
[
"Ma",
"H. L.",
""
],
[
"Ma",
"J. L.",
""
],
[
"Ma",
"L. L.",
""
],
[
"Ma",
"L. R.",
""
],
[
"Ma",
"Q. M.",
""
],
[
"Ma",
"R. Q.",
""
],
[
"Ma",
"R. Y.",
""
],
[
"Ma",
"T.",
""
],
[
"Ma",
"X. T.",
""
],
[
"Ma",
"X. Y.",
""
],
[
"Ma",
"Y. M.",
""
],
[
"Maas",
"F. E.",
""
],
[
"MacKay",
"I.",
""
],
[
"Maggiora",
"M.",
""
],
[
"Malde",
"S.",
""
],
[
"Mao",
"Y. J.",
""
],
[
"Mao",
"Z. P.",
""
],
[
"Marcello",
"S.",
""
],
[
"Melendi",
"F. M.",
""
],
[
"Meng",
"Y. H.",
""
],
[
"Meng",
"Z. X.",
""
],
[
"Messchendorp",
"J. G.",
""
],
[
"Mezzadri",
"G.",
""
],
[
"Miao",
"H.",
""
],
[
"Min",
"T. J.",
""
],
[
"Mitchell",
"R. E.",
""
],
[
"Mo",
"X. H.",
""
],
[
"Moses",
"B.",
""
],
[
"Muchnoi",
"N. Yu.",
""
],
[
"Muskalla",
"J.",
""
],
[
"Nefedov",
"Y.",
""
],
[
"Nerling",
"F.",
""
],
[
"Nie",
"L. S.",
""
],
[
"Nikolaev",
"I. B.",
""
],
[
"Ning",
"Z.",
""
],
[
"Nisar",
"S.",
""
],
[
"Niu",
"Q. L.",
""
],
[
"Niu",
"W. D.",
""
],
[
"Olsen",
"S. L.",
""
],
[
"Ouyang",
"Q.",
""
],
[
"Pacetti",
"S.",
""
],
[
"Pan",
"X.",
""
],
[
"Pan",
"Y.",
""
],
[
"Pathak",
"A.",
""
],
[
"Pei",
"Y. P.",
""
],
[
"Pelizaeus",
"M.",
""
],
[
"Peng",
"H. P.",
""
],
[
"Peng",
"Y. Y.",
""
],
[
"Peters",
"K.",
""
],
[
"Ping",
"J. L.",
""
],
[
"Ping",
"R. G.",
""
],
[
"Plura",
"S.",
""
],
[
"Prasad",
"V.",
""
],
[
"Qi",
"F. Z.",
""
],
[
"Qi",
"H. R.",
""
],
[
"Qi",
"M.",
""
],
[
"Qian",
"S.",
""
],
[
"Qian",
"W. B.",
""
],
[
"Qiao",
"C. F.",
""
],
[
"Qiao",
"J. H.",
""
],
[
"Qin",
"J. J.",
""
],
[
"Qin",
"J. L.",
""
],
[
"Qin",
"L. Q.",
""
],
[
"Qin",
"L. Y.",
""
],
[
"Qin",
"P. B.",
""
],
[
"Qin",
"X. P.",
""
],
[
"Qin",
"X. S.",
""
],
[
"Qin",
"Z. H.",
""
],
[
"Qiu",
"J. F.",
""
],
[
"Qu",
"Z. H.",
""
],
[
"Redmer",
"C. F.",
""
],
[
"Rivetti",
"A.",
""
],
[
"Rolo",
"M.",
""
],
[
"Rong",
"G.",
""
],
[
"Rong",
"S. S.",
""
],
[
"Rosini",
"F.",
""
],
[
"Rosner",
"Ch.",
""
],
[
"Ruan",
"M. Q.",
""
],
[
"Salone",
"N.",
""
],
[
"Sarantsev",
"A.",
""
],
[
"Schelhaas",
"Y.",
""
],
[
"Schoenning",
"K.",
""
],
[
"Scodeggio",
"M.",
""
],
[
"Shan",
"K. Y.",
""
],
[
"Shan",
"W.",
""
],
[
"Shan",
"X. Y.",
""
],
[
"Shang",
"Z. J.",
""
],
[
"Shangguan",
"J. F.",
""
],
[
"Shao",
"L. G.",
""
],
[
"Shao",
"M.",
""
],
[
"Shen",
"C. P.",
""
],
[
"Shen",
"H. F.",
""
],
[
"Shen",
"W. H.",
""
],
[
"Shen",
"X. Y.",
""
],
[
"Shi",
"B. A.",
""
],
[
"Shi",
"H.",
""
],
[
"Shi",
"J. L.",
""
],
[
"Shi",
"J. Y.",
""
],
[
"Shi",
"S. Y.",
""
],
[
"Shi",
"X.",
""
],
[
"Song",
"H. L.",
""
],
[
"Song",
"J. J.",
""
],
[
"Song",
"T. Z.",
""
],
[
"Song",
"W. M.",
""
],
[
"Song",
"Y. X.",
""
],
[
"Sosio",
"S.",
""
],
[
"Spataro",
"S.",
""
],
[
"Stieler",
"F.",
""
],
[
"Su",
"S. S",
""
],
[
"Su",
"Y. J.",
""
],
[
"Sun",
"G. B.",
""
],
[
"Sun",
"G. X.",
""
],
[
"Sun",
"H.",
""
],
[
"Sun",
"H. K.",
""
],
[
"Sun",
"J. F.",
""
],
[
"Sun",
"K.",
""
],
[
"Sun",
"L.",
""
],
[
"Sun",
"S. S.",
""
],
[
"Sun",
"T.",
""
],
[
"Sun",
"Y. C.",
""
],
[
"Sun",
"Y. H.",
""
],
[
"Sun",
"Y. J.",
""
],
[
"Sun",
"Y. Z.",
""
],
[
"Sun",
"Z. Q.",
""
],
[
"Sun",
"Z. T.",
""
],
[
"Tang",
"C. J.",
""
],
[
"Tang",
"G. Y.",
""
],
[
"Tang",
"J.",
""
],
[
"Tang",
"L. F.",
""
],
[
"Tang",
"M.",
""
],
[
"Tang",
"Y. A.",
""
],
[
"Tao",
"L. Y.",
""
],
[
"Tat",
"M.",
""
],
[
"Teng",
"J. X.",
""
],
[
"Tian",
"J. Y.",
""
],
[
"Tian",
"W. H.",
""
],
[
"Tian",
"Y.",
""
],
[
"Tian",
"Z. F.",
""
],
[
"Uman",
"I.",
""
],
[
"Wang",
"B.",
""
],
[
"Wang",
"B.",
""
],
[
"Wang",
"Bo",
""
],
[
"Wang",
"C.",
""
],
[
"Wang",
"Cong",
""
],
[
"Wang",
"D. Y.",
""
],
[
"Wang",
"H. J.",
""
],
[
"Wang",
"J. J.",
""
],
[
"Wang",
"K.",
""
],
[
"Wang",
"L. L.",
""
],
[
"Wang",
"L. W.",
""
],
[
"Wang",
"M.",
""
],
[
"Wang",
"M.",
""
],
[
"Wang",
"N. Y.",
""
],
[
"Wang",
"S.",
""
],
[
"Wang",
"T.",
""
],
[
"Wang",
"T. J.",
""
],
[
"Wang",
"W.",
""
],
[
"Wang",
"W.",
""
],
[
"Wang",
"W. P.",
""
],
[
"Wang",
"X.",
""
],
[
"Wang",
"X. F.",
""
],
[
"Wang",
"X. J.",
""
],
[
"Wang",
"X. L.",
""
],
[
"Wang",
"X. N.",
""
],
[
"Wang",
"Y.",
""
],
[
"Wang",
"Y. D.",
""
],
[
"Wang",
"Y. F.",
""
],
[
"Wang",
"Y. H.",
""
],
[
"Wang",
"Y. L.",
""
],
[
"Wang",
"Y. N.",
""
],
[
"Wang",
"Y. Q.",
""
],
[
"Wang",
"Yaqian",
""
],
[
"Wang",
"Yi",
""
],
[
"Wang",
"Yuan",
""
],
[
"Wang",
"Z.",
""
],
[
"Wang",
"Z. L.",
""
],
[
"Wang",
"Z. L.",
""
],
[
"Wang",
"Z. Q.",
""
],
[
"Wang",
"Z. Y.",
""
],
[
"Wei",
"D. H.",
""
],
[
"Wei",
"H. R.",
""
],
[
"Weidner",
"F.",
""
],
[
"Wen",
"S. P.",
""
],
[
"Wen",
"Y. R.",
""
],
[
"Wiedner",
"U.",
""
],
[
"Wilkinson",
"G.",
""
],
[
"Wolke",
"M.",
""
],
[
"Wu",
"C.",
""
],
[
"Wu",
"J. F.",
""
],
[
"Wu",
"L. H.",
""
],
[
"Wu",
"L. J.",
""
],
[
"Wu",
"L. J.",
""
],
[
"Wu",
"Lianjie",
""
],
[
"Wu",
"S. G.",
""
],
[
"Wu",
"S. M.",
""
],
[
"Wu",
"X.",
""
],
[
"Wu",
"X. H.",
""
],
[
"Wu",
"Y. J.",
""
],
[
"Wu",
"Z.",
""
],
[
"Xia",
"L.",
""
],
[
"Xian",
"X. M.",
""
],
[
"Xiang",
"B. H.",
""
],
[
"Xiao",
"D.",
""
],
[
"Xiao",
"G. Y.",
""
],
[
"Xiao",
"H.",
""
],
[
"Xiao",
"Y. L.",
""
],
[
"Xiao",
"Z. J.",
""
],
[
"Xie",
"C.",
""
],
[
"Xie",
"K. J.",
""
],
[
"Xie",
"X. H.",
""
],
[
"Xie",
"Y.",
""
],
[
"Xie",
"Y. G.",
""
],
[
"Xie",
"Y. H.",
""
],
[
"Xie",
"Z. P.",
""
],
[
"Xing",
"T. Y.",
""
],
[
"Xu",
"C. F.",
""
],
[
"Xu",
"C. J.",
""
],
[
"Xu",
"G. F.",
""
],
[
"Xu",
"H. Y.",
""
],
[
"Xu",
"H. Y.",
""
],
[
"Xu",
"M.",
""
],
[
"Xu",
"Q. J.",
""
],
[
"Xu",
"Q. N.",
""
],
[
"Xu",
"T. D.",
""
],
[
"Xu",
"W. L.",
""
],
[
"Xu",
"X. P.",
""
],
[
"Xu",
"Y.",
""
],
[
"Xu",
"Y.",
""
],
[
"Xu",
"Y. C.",
""
],
[
"Xu",
"Z. S.",
""
],
[
"Yan",
"H. Y.",
""
],
[
"Yan",
"L.",
""
],
[
"Yan",
"W. B.",
""
],
[
"Yan",
"W. C.",
""
],
[
"Yan",
"W. P.",
""
],
[
"Yan",
"X. Q.",
""
],
[
"Yang",
"H. J.",
""
],
[
"Yang",
"H. L.",
""
],
[
"Yang",
"H. X.",
""
],
[
"Yang",
"J. H.",
""
],
[
"Yang",
"R. J.",
""
],
[
"Yang",
"T.",
""
],
[
"Yang",
"Y.",
""
],
[
"Yang",
"Y. F.",
""
],
[
"Yang",
"Y. H.",
""
],
[
"Yang",
"Y. Q.",
""
],
[
"Yang",
"Y. X.",
""
],
[
"Yang",
"Y. Z.",
""
],
[
"Ye",
"M.",
""
],
[
"Ye",
"M. H.",
""
],
[
"Ye",
"Z. J.",
""
],
[
"Yin",
"Junhao",
""
],
[
"You",
"Z. Y.",
""
],
[
"Yu",
"B. X.",
""
],
[
"Yu",
"C. X.",
""
],
[
"Yu",
"G.",
""
],
[
"Yu",
"J. S.",
""
],
[
"Yu",
"M. C.",
""
],
[
"Yu",
"T.",
""
],
[
"Yu",
"X. D.",
""
],
[
"Yu",
"Y. C.",
""
],
[
"Yuan",
"C. Z.",
""
],
[
"Yuan",
"H.",
""
],
[
"Yuan",
"J.",
""
],
[
"Yuan",
"J.",
""
],
[
"Yuan",
"L.",
""
],
[
"Yuan",
"S. C.",
""
],
[
"Yuan",
"Y.",
""
],
[
"Yuan",
"Z. Y.",
""
],
[
"Yue",
"C. X.",
""
],
[
"Yue",
"Ying",
""
],
[
"Zafar",
"A. A.",
""
],
[
"Zeng",
"S. H.",
""
],
[
"Zeng",
"X.",
""
],
[
"Zeng",
"Y.",
""
],
[
"Zeng",
"Y. J.",
""
],
[
"Zeng",
"Y. J.",
""
],
[
"Zhai",
"X. Y.",
""
],
[
"Zhan",
"Y. H.",
""
],
[
"Zhang",
"A. Q.",
""
],
[
"Zhang",
"B. L.",
""
],
[
"Zhang",
"B. X.",
""
],
[
"Zhang",
"D. H.",
""
],
[
"Zhang",
"G. Y.",
""
],
[
"Zhang",
"G. Y.",
""
],
[
"Zhang",
"H.",
""
],
[
"Zhang",
"H.",
""
],
[
"Zhang",
"H. C.",
""
],
[
"Zhang",
"H. H.",
""
],
[
"Zhang",
"H. Q.",
""
],
[
"Zhang",
"H. R.",
""
],
[
"Zhang",
"H. Y.",
""
],
[
"Zhang",
"J.",
""
],
[
"Zhang",
"J.",
""
],
[
"Zhang",
"J. J.",
""
],
[
"Zhang",
"J. L.",
""
],
[
"Zhang",
"J. Q.",
""
],
[
"Zhang",
"J. S.",
""
],
[
"Zhang",
"J. W.",
""
],
[
"Zhang",
"J. X.",
""
],
[
"Zhang",
"J. Y.",
""
],
[
"Zhang",
"J. Z.",
""
],
[
"Zhang",
"Jianyu",
""
],
[
"Zhang",
"L. M.",
""
],
[
"Zhang",
"Lei",
""
],
[
"Zhang",
"N.",
""
],
[
"Zhang",
"P.",
""
],
[
"Zhang",
"Q.",
""
],
[
"Zhang",
"Q. Y.",
""
],
[
"Zhang",
"R. Y.",
""
],
[
"Zhang",
"S. H.",
""
],
[
"Zhang",
"Shulei",
""
],
[
"Zhang",
"X. M.",
""
],
[
"Zhang",
"X. Y",
""
],
[
"Zhang",
"X. Y.",
""
],
[
"Zhang",
"Y.",
""
],
[
"Zhang",
"Y.",
""
],
[
"Zhang",
"Y. T.",
""
],
[
"Zhang",
"Y. H.",
""
],
[
"Zhang",
"Y. M.",
""
],
[
"Zhang",
"Z. D.",
""
],
[
"Zhang",
"Z. H.",
""
],
[
"Zhang",
"Z. L.",
""
],
[
"Zhang",
"Z. L.",
""
],
[
"Zhang",
"Z. X.",
""
],
[
"Zhang",
"Z. Y.",
""
],
[
"Zhang",
"Z. Y.",
""
],
[
"Zhang",
"Z. Z.",
""
],
[
"Zhang",
"Zh. Zh.",
""
],
[
"Zhao",
"G.",
""
],
[
"Zhao",
"J. Y.",
""
],
[
"Zhao",
"J. Z.",
""
],
[
"Zhao",
"L.",
""
],
[
"Zhao",
"Lei",
""
],
[
"Zhao",
"M. G.",
""
],
[
"Zhao",
"N.",
""
],
[
"Zhao",
"R. P.",
""
],
[
"Zhao",
"S. J.",
""
],
[
"Zhao",
"Y. B.",
""
],
[
"Zhao",
"Y. L.",
""
],
[
"Zhao",
"Y. X.",
""
],
[
"Zhao",
"Z. G.",
""
],
[
"Zhemchugov",
"A.",
""
],
[
"Zheng",
"B.",
""
],
[
"Zheng",
"B. M.",
""
],
[
"Zheng",
"J. P.",
""
],
[
"Zheng",
"W. J.",
""
],
[
"Zheng",
"X. R.",
""
],
[
"Zheng",
"Y. H.",
""
],
[
"Zhong",
"B.",
""
],
[
"Zhong",
"C.",
""
],
[
"Zhou",
"H.",
""
],
[
"Zhou",
"J. Q.",
""
],
[
"Zhou",
"J. Y.",
""
],
[
"Zhou",
"S.",
""
],
[
"Zhou",
"X.",
""
],
[
"Zhou",
"X. K.",
""
],
[
"Zhou",
"X. R.",
""
],
[
"Zhou",
"X. Y.",
""
],
[
"Zhou",
"Y. Z.",
""
],
[
"Zhou",
"Z. C.",
""
],
[
"Zhu",
"A. N.",
""
],
[
"Zhu",
"J.",
""
],
[
"Zhu",
"K.",
""
],
[
"Zhu",
"K. J.",
""
],
[
"Zhu",
"K. S.",
""
],
[
"Zhu",
"L.",
""
],
[
"Zhu",
"L. X.",
""
],
[
"Zhu",
"S. H.",
""
],
[
"Zhu",
"T. J.",
""
],
[
"Zhu",
"W. D.",
""
],
[
"Zhu",
"W. D.",
""
],
[
"Zhu",
"W. J.",
""
],
[
"Zhu",
"W. Z.",
""
],
[
"Zhu",
"Y. C.",
""
],
[
"Zhu",
"Z. A.",
""
],
[
"Zhuang",
"X. Y.",
""
],
[
"Zou",
"J. H.",
""
],
[
"Zu",
"J.",
""
]
] |
2506.02522 | Xu Wan | Xu Wan, Wenyue Xu, Chao Yang, Mingyang Sun | Think Twice, Act Once: A Co-Evolution Framework of LLM and RL for Large-Scale Decision Making | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Recent advancements in Large Language Models (LLMs) and Reinforcement Learning (RL) have shown significant promise in decision-making tasks. Nevertheless, for large-scale industrial decision problems, both approaches face distinct challenges: LLMs lack real-time long-sequence decision-making capabilities, while RL struggles with sample efficiency in vast action spaces. To bridge this gap, we propose Agents Co-Evolution (ACE), a synergistic framework between LLMs and RL agents for large-scale decision-making scenarios. ACE introduces a dual-role trajectory refinement mechanism where LLMs act as both Policy Actor and Value Critic during RL's training: the Actor refines suboptimal actions via multi-step reasoning and environment validation, while the Critic performs temporal credit assignment through trajectory-level reward shaping. Concurrently, RL agent enhances LLMs' task-specific decision-making with high-quality fine-tuning datasets generated via prioritized experience replay. Through extensive experiments across multiple power grid operation challenges with action spaces exceeding 60K discrete actions, ACE demonstrates superior performance over existing RL methods and LLM-based methods. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:52:37 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Wan",
"Xu",
""
],
[
"Xu",
"Wenyue",
""
],
[
"Yang",
"Chao",
""
],
[
"Sun",
"Mingyang",
""
]
] |
2506.02523 | Robin Geens | Robin Geens, Marian Verhelst | Hardware-Centric Analysis of DeepSeek's Multi-Head Latent Attention | null | null | null | null | cs.AR | http://creativecommons.org/licenses/by/4.0/ | Multi-Head Latent Attention (MLA), introduced in DeepSeek-V2, improves the efficiency of large language models by projecting query, key, and value tensors into a compact latent space. This architectural change reduces the KV-cache size and significantly lowers memory bandwidth demands, particularly in the autoregressive decode phase. This letter presents the first hardware-centric analysis of MLA, comparing it to conventional Multi-Head Attention (MHA) and evaluating its implications for accelerator performance. We identify two alternative execution schemes of MLA--reusing, resp. recomputing latent projection matrices--which offer distinct trade-offs between compute and memory access. Using the Stream design space exploration framework, we model their throughput and energy cost across a range of hardware platforms and find that MLA can shift attention workloads toward the compute-bound regime.
Our results show that MLA not only reduces bandwidth usage but also enables adaptable execution strategies aligned with hardware constraints. Compared to MHA, it provides more stable and efficient performance, particularly on bandwidth-limited hardware platforms. These findings emphasize MLA's relevance as a co-design opportunity for future AI accelerators. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:53:04 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Geens",
"Robin",
""
],
[
"Verhelst",
"Marian",
""
]
] |
2506.02524 | Rahul Ghosal | Yuanzhen Yue, Stella Self, Yichao Wu, Jiajia Zhang, Rahul Ghosal | Variable Selection in Functional Linear Cox Model | null | null | null | null | stat.ME stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern biomedical studies frequently collect complex, high-dimensional physiological signals using wearables and sensors along with time-to-event outcomes, making efficient variable selection methods crucial for interpretation and improving the accuracy of survival models. We propose a novel variable selection method for a functional linear Cox model with multiple functional and scalar covariates measured at baseline. We utilize a spline-based semiparametric estimation approach for the functional coefficients and a group minimax concave type penalty (MCP), which effectively integrates smoothness and sparsity into the estimation of functional coefficients. An efficient group descent algorithm is used for optimization, and an automated procedure is provided to select optimal values of the smoothing and sparsity parameters. Through simulation studies, we demonstrate the method's ability to perform accurate variable selection and estimation. The method is applied to 2003-06 cohort of the National Health and Nutrition Examination Survey (NHANES) data, identifying the key temporally varying distributional patterns of physical activity and demographic predictors related to all-cause mortality. Our analysis sheds light on the intricate association between daily distributional patterns of physical activity and all-cause mortality among older US adults. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:56:17 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Yue",
"Yuanzhen",
""
],
[
"Self",
"Stella",
""
],
[
"Wu",
"Yichao",
""
],
[
"Zhang",
"Jiajia",
""
],
[
"Ghosal",
"Rahul",
""
]
] |
2506.02525 | Alonso Espinoza Rojas | Alonso Espinoza, Eric Goles and Marco Montalva-Medel | Boolean-network simplification and rule fitting to unravel chemotherapy resistance in non-small cell lung cancer | null | null | null | null | cs.DM | http://creativecommons.org/publicdomain/zero/1.0/ | Boolean networks are powerful frameworks for capturing the logic of gene-regulatory circuits, yet their combinatorial explosion hampers exhaustive analyses. Here, we present a systematic reduction of a 31-node Boolean model that describes cisplatin- and pemetrexed-resistance in non-small-cell lung cancer to a compact 9-node core that exactly reproduces the original attractor landscape. The streamlined network shrinks the state space by four orders of magnitude, enabling rapid exploration of critical control points, rules fitting and candidate therapeutic targets. Extensive synchronous and asynchronous simulations confirm that the three clinically relevant steady states and their basins of attraction are conserved and reflect resistance frequencies close to those reported in clinical studies. The reduced model provides an accessible scaffold for future mechanistic and drug-discovery studies. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:56:26 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Espinoza",
"Alonso",
""
],
[
"Goles",
"Eric",
""
],
[
"Montalva-Medel",
"Marco",
""
]
] |
2506.02526 | Wei Hao | Wei Hao, Jia-Jun Wu, Jin-Lin Fu | The Study of Pole Trajectory within a bare state in the coupled channel model | null | null | null | null | hep-ph nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate two-particle scattering and two-particle scattering with a bare basis state using Hamiltonian Effective Field Theory (HEFT). We analyze the distribution of two-body scattering poles in the momentum and energy planes under relativistic conditions. Compared to the non-relativistic case, there are significant differences in the distribution of bound state poles and resonance poles in the relativistic case, primarily due to the square root term in the relativistic formula. By considering pure two-particle scattering, we examine the relationship between the form factor and the number of poles. Additionally, we clearly elucidate the effects of attractive and repulsive interactions on the bound state poles and resonance poles. More importantly, we extend our model by including a bare state and explore the poles originating from the bare state or coupled channels through the trajectories of pole positions, as well as the compositeness of bound states. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 06:59:39 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Hao",
"Wei",
""
],
[
"Wu",
"Jia-Jun",
""
],
[
"Fu",
"Jin-Lin",
""
]
] |
2506.02527 | Yingying Zhuang Dr | Yingying Zhuang, Aman Gupta, Anurag Beniwal | Multilingual Information Retrieval with a Monolingual Knowledge Base | 6 pages, accepted at GENNEXT@SIGIR25 | null | null | null | cs.CL cs.AI cs.IR | http://creativecommons.org/licenses/by/4.0/ | Multilingual information retrieval has emerged as powerful tools for expanding knowledge sharing across languages. On the other hand, resources on high quality knowledge base are often scarce and in limited languages, therefore an effective embedding model to transform sentences from different languages into a feature vector space same as the knowledge base language becomes the key ingredient for cross language knowledge sharing, especially to transfer knowledge available in high-resource languages to low-resource ones. In this paper we propose a novel strategy to fine-tune multilingual embedding models with weighted sampling for contrastive learning, enabling multilingual information retrieval with a monolingual knowledge base. We demonstrate that the weighted sampling strategy produces performance gains compared to standard ones by up to 31.03\% in MRR and up to 33.98\% in Recall@3. Additionally, our proposed methodology is language agnostic and applicable for both multilingual and code switching use cases. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:05:49 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zhuang",
"Yingying",
""
],
[
"Gupta",
"Aman",
""
],
[
"Beniwal",
"Anurag",
""
]
] |
2506.02528 | Yiren Song | Yan Gong, Yiren Song, Yicheng Li, Chenglin Li, Yin Zhang | RelationAdapter: Learning and Transferring Visual Relation with Diffusion Transformers | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Inspired by the in-context learning mechanism of large language models (LLMs), a new paradigm of generalizable visual prompt-based image editing is emerging. Existing single-reference methods typically focus on style or appearance adjustments and struggle with non-rigid transformations. To address these limitations, we propose leveraging source-target image pairs to extract and transfer content-aware editing intent to novel query images. To this end, we introduce RelationAdapter, a lightweight module that enables Diffusion Transformer (DiT) based models to effectively capture and apply visual transformations from minimal examples. We also introduce Relation252K, a comprehensive dataset comprising 218 diverse editing tasks, to evaluate model generalization and adaptability in visual prompt-driven scenarios. Experiments on Relation252K show that RelationAdapter significantly improves the model's ability to understand and transfer editing intent, leading to notable gains in generation quality and overall editing performance. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:06:35 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Gong",
"Yan",
""
],
[
"Song",
"Yiren",
""
],
[
"Li",
"Yicheng",
""
],
[
"Li",
"Chenglin",
""
],
[
"Zhang",
"Yin",
""
]
] |
2506.02529 | Khang Le | Nguyen-Khang Le, Quan Minh Bui, Minh Ngoc Nguyen, Hiep Nguyen, Trung Vo, Son T. Luu, Shoshin Nomura, Minh Le Nguyen | Automated Web Application Testing: End-to-End Test Case Generation with Large Language Models and Screen Transition Graphs | Published in the Proceedings of JSAI 2025 | null | null | null | cs.SE cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Web applications are critical to modern software ecosystems, yet ensuring their reliability remains challenging due to the complexity and dynamic nature of web interfaces. Recent advances in large language models (LLMs) have shown promise in automating complex tasks, but limitations persist in handling dynamic navigation flows and complex form interactions. This paper presents an automated system for generating test cases for two key aspects of web application testing: site navigation and form filling. For site navigation, the system employs screen transition graphs and LLMs to model navigation flows and generate test scenarios. For form filling, it uses state graphs to handle conditional forms and automates Selenium script generation. Key contributions include: (1) a novel integration of graph structures and LLMs for site navigation testing, (2) a state graph-based approach for automating form-filling test cases, and (3) a comprehensive dataset for evaluating form-interaction testing. Experimental results demonstrate the system's effectiveness in improving test coverage and robustness, advancing the state of web application testing. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:08:21 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Le",
"Nguyen-Khang",
""
],
[
"Bui",
"Quan Minh",
""
],
[
"Nguyen",
"Minh Ngoc",
""
],
[
"Nguyen",
"Hiep",
""
],
[
"Vo",
"Trung",
""
],
[
"Luu",
"Son T.",
""
],
[
"Nomura",
"Shoshin",
""
],
[
"Nguyen",
"Minh Le",
""
]
] |
2506.02530 | Sho Kubota | Sho Kubota, Hiroto Sekido, Harunobu Yata, Kiyoto Yoshino | Strongly regular and strongly walk-regular graphs that admit perfect state transfer | 21 pages, | null | null | null | math.CO quant-ph | http://creativecommons.org/publicdomain/zero/1.0/ | We study perfect state transfer in Grover walks on two important classes of graphs: strongly regular graphs and strongly walk-regular graphs. The latter class is a generalization of the former. We first give a complete classification of strongly regular graphs that admit perfect state transfer. The only such graphs are the complete bipartite graph $K_{2,2}$ and the complete tripartite graph $K_{2,2,2}$. We then show that, if a genuine strongly walk-regular graph admits perfect state transfer, then its spectrum must be of the form $\{[k]^1, [\frac{k}{2}]^{\alpha}, [0]^{\beta}, [-\frac{k}{2}]^{\gamma}\}$, and we enumerate all feasible spectra of this form up to $k=20$ with the help of a computer. These results are obtained using techniques from algebraic number theory and spectral graph theory, particularly through the analysis of eigenvalues and eigenprojections of a normalized adjacency matrix. While the setting is in quantum walks, the core discussion is developed entirely within the framework of spectral graph theory. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:10:06 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kubota",
"Sho",
""
],
[
"Sekido",
"Hiroto",
""
],
[
"Yata",
"Harunobu",
""
],
[
"Yoshino",
"Kiyoto",
""
]
] |
2506.02531 | Preeti Bhall | Preeti Bhall, Ritu Garg and Alka Upadhyay | Impact of sea quark-gluon configurations on the magnetic octupole deformation of decuplet baryons | null | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | The magnetic octupole moment of $J^P= \frac{3}{2}^+$ decuplet baryons are discussed in the statistical framework, treating baryons as an ensembles of quark-gluon Fock states. The probabilities associated with multiple strange and non-strange Fock states depict the importance of sea in spin, flavor $\&$ color space, which are further merged into statistical parameters. The individual contribution of valence and sea (scalar, vector and tensor) to the magnetic octupole moment is calculated. The symmetry breaking in both sea and valence is experienced by a suppression factor $k(1-C_l)^{n-1}$ and a mass correction parameter 'r', respectively. The factor $k(1-C_l)^{n-1}$ systematically reduces the probabilities of Fock states containing multiple strange quark pairs. The octupole moment value is obtained -ve for $\Delta^{++}, \Delta^+, \Sigma^{*+}$ and +ve for $\Delta^{-}, \Sigma^{*-}, \Xi^{*-}, \Omega^-$ baryons with the domination of scalar (spin-0) sea. The computed results are compared with existing theoretical predictions, demonstrating good consistency. These predictions may serve as valuable inputs for future high-precision experiments and theoretical explorations in hadron structure. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:10:34 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Bhall",
"Preeti",
""
],
[
"Garg",
"Ritu",
""
],
[
"Upadhyay",
"Alka",
""
]
] |
2506.02532 | Jinu Lee | Jinu Lee, Sagnik Mukherjee, Dilek Hakkani-Tur, Julia Hockenmaier | ReasoningFlow: Semantic Structure of Complex Reasoning Traces | 10 pages, 6 figures. ArgMining 2025 Workshop (Non-archival) @ ACL 2025 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large reasoning models (LRMs) generate complex reasoning traces with planning, reflection, verification, and backtracking. In this work, we introduce ReasoningFlow, a unified schema for analyzing the semantic structures of these complex traces. ReasoningFlow parses traces into directed acyclic graphs, enabling the characterization of distinct reasoning patterns as subgraph structures. This human-interpretable representation offers promising applications in understanding, evaluating, and enhancing the reasoning processes of LRMs. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:11:34 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Lee",
"Jinu",
""
],
[
"Mukherjee",
"Sagnik",
""
],
[
"Hakkani-Tur",
"Dilek",
""
],
[
"Hockenmaier",
"Julia",
""
]
] |
2506.02533 | Maike Behrendt | Maike Behrendt, Stefan Sylvius Wagner, Carina Weinmann, Marike Bormann, Mira Warne, Stefan Harmeling | Natural Language Processing to Enhance Deliberation in Political Online Discussions: A Survey | null | null | null | null | cs.CL cs.HC | http://creativecommons.org/licenses/by/4.0/ | Political online participation in the form of discussing political issues and exchanging opinions among citizens is gaining importance with more and more formats being held digitally. To come to a decision, a careful discussion and consideration of opinions and a civil exchange of arguments, which is defined as the act of deliberation, is desirable. The quality of discussions and participation processes in terms of their deliberativeness highly depends on the design of platforms and processes. To facilitate online communication for both participants and initiators, machine learning methods offer a lot of potential. In this work we want to showcase which issues occur in political online discussions and how machine learning can be used to counteract these issues and enhance deliberation. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:11:49 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Behrendt",
"Maike",
""
],
[
"Wagner",
"Stefan Sylvius",
""
],
[
"Weinmann",
"Carina",
""
],
[
"Bormann",
"Marike",
""
],
[
"Warne",
"Mira",
""
],
[
"Harmeling",
"Stefan",
""
]
] |
2506.02534 | Sining Chen | Sining Chen, Yilei Shi, Xiao Xiang Zhu | Enhancing Monocular Height Estimation via Weak Supervision from Imperfect Labels | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Monocular height estimation is considered the most efficient and cost-effective means of 3D perception in remote sensing, and it has attracted much attention since the emergence of deep learning. While training neural networks requires a large amount of data, data with perfect labels are scarce and only available within developed regions. The trained models therefore lack generalizability, which limits the potential for large-scale application of existing methods. We tackle this problem for the first time, by introducing data with imperfect labels into training pixel-wise height estimation networks, including labels that are incomplete, inexact, and inaccurate compared to high-quality labels. We propose an ensemble-based pipeline compatible with any monocular height estimation network. Taking the challenges of noisy labels, domain shift, and long-tailed distribution of height values into consideration, we carefully design the architecture and loss functions to leverage the information concealed in imperfect labels using weak supervision through balanced soft losses and ordinal constraints. We conduct extensive experiments on two datasets with different resolutions, DFC23 (0.5 to 1 m) and GBH (3 m). The results indicate that the proposed pipeline outperforms baselines by achieving more balanced performance across various domains, leading to improvements of average root mean square errors up to 22.94 %, and 18.62 % on DFC23 and GBH, respectively. The efficacy of each design component is validated through ablation studies. Code is available at https://github.com/zhu-xlab/weakim2h. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:14:16 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Chen",
"Sining",
""
],
[
"Shi",
"Yilei",
""
],
[
"Zhu",
"Xiao Xiang",
""
]
] |
2506.02536 | Xin Liu | Xin Liu, Lu Wang | Answer Convergence as a Signal for Early Stopping in Reasoning | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Chain-of-thought (CoT) prompting enhances reasoning in large language models (LLMs) but often leads to verbose and redundant outputs, thus increasing inference cost. We hypothesize that many reasoning steps are unnecessary for producing correct answers. To investigate this, we start with a systematic study to examine what is the minimum reasoning required for a model to reach a stable decision. We find that on math reasoning tasks like math, models typically converge to their final answers after 60\% of the reasoning steps, suggesting substantial redundancy in the remaining content. Based on these insights, we propose three inference-time strategies to improve efficiency: (1) early stopping via answer consistency, (2) boosting the probability of generating end-of-reasoning signals, and (3) a supervised method that learns when to stop based on internal activations. Experiments across five benchmarks and five open-weights LLMs show that our methods significantly reduce token usage with little or no accuracy drop. In particular, on NaturalQuestions, Answer Consistency reduces tokens by over 40\% while further improving accuracy. Our work underscores the importance of cost-effective reasoning methods that operate at inference time, offering practical benefits for real-world applications. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:20:54 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Liu",
"Xin",
""
],
[
"Wang",
"Lu",
""
]
] |
2506.02537 | Hao Yan | Hao Yan, Handong Zheng, Hao Wang, Liang Yin, Xingchen Liu, Zhenbiao Cao, Xinxing Su, Zihao Chen, Jihao Wu, Minghui Liao, Chao Weng, Wei Chen, Yuliang Liu, Xiang Bai | VisuRiddles: Fine-grained Perception is a Primary Bottleneck for Multimodal Large Language Models in Abstract Visual Reasoning | 13 pages, 4 figures | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Recent strides in multimodal large language models (MLLMs) have significantly advanced their performance in many reasoning tasks. However, Abstract Visual Reasoning (AVR) remains a critical challenge, primarily due to limitations in perceiving abstract graphics. To tackle this issue, we investigate the bottlenecks in current MLLMs and synthesize training data to improve their abstract visual perception. First, we propose VisuRiddles, a benchmark for AVR, featuring tasks meticulously constructed to assess models' reasoning capacities across five core dimensions and two high-level reasoning categories. Second, we introduce the Perceptual Riddle Synthesizer (PRS), an automated framework for generating riddles with fine-grained perceptual descriptions. PRS not only generates valuable training data for abstract graphics but also provides fine-grained perceptual description, crucially allowing for supervision over intermediate reasoning stages and thereby improving both training efficacy and model interpretability. Our extensive experimental results on VisuRiddles empirically validate that fine-grained visual perception is the principal bottleneck and our synthesis framework markedly enhances the performance of contemporary MLLMs on these challenging tasks. Our code and dataset will be released at https://github.com/yh-hust/VisuRiddles | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:24:00 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Yan",
"Hao",
""
],
[
"Zheng",
"Handong",
""
],
[
"Wang",
"Hao",
""
],
[
"Yin",
"Liang",
""
],
[
"Liu",
"Xingchen",
""
],
[
"Cao",
"Zhenbiao",
""
],
[
"Su",
"Xinxing",
""
],
[
"Chen",
"Zihao",
""
],
[
"Wu",
"Jihao",
""
],
[
"Liao",
"Minghui",
""
],
[
"Weng",
"Chao",
""
],
[
"Chen",
"Wei",
""
],
[
"Liu",
"Yuliang",
""
],
[
"Bai",
"Xiang",
""
]
] |
2506.02538 | Emilio Mart\'inez-Pa\~neda | C. Cui, L. Cupertino-Malheiros, Z. Xiong, E. Mart\'inez-Pa\~neda | On the fracture mechanics validity of small scale tests | null | null | null | null | cs.CE physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There is growing interest in conducting small-scale tests to gain additional insight into the fracture behaviour of components across a wide range of materials. For example, micro-scale mechanical tests inside of a microscope (\emph{in situ}) enable direct, high-resolution observation of the interplay between crack growth and microstructural phenomena (e.g., dislocation behaviour or the fracture resistance of a particular interface), and sub-size samples are increasingly used when only a limited amount of material is available. However, to obtain quantitative insight and extract relevant fracture parameters, the sample must be sufficiently large for a $J$- (HRR) or a $K$-field to exist. We conduct numerical and semi-analytical studies to map the conditions (sample geometry, material) that result in a valid, quantitative fracture experiment. Specifically, for a wide range of material properties, crack lengths and sample dimensions, we establish the maximum value of the $J$-integral where an HRR field ceases to exist (i.e., the maximum $J$ value at which fracture must occur for the test to be valid, $J_\mathrm{max}$). Maps are generated to establish the maximum valid $J$ value ($J_\mathrm{max}$) as a function of yield strength, strain hardening and minimum sample size. These maps are then used to discuss the existing experimental literature and provide guidance on how to conduct quantitative experiments. Finally, our study is particularised to the analysis of metals that have been embrittled due to hydrogen exposure. The response of relevant materials under hydrogen-containing environments are superimposed on the aforementioned maps, determining the conditions that will enable quantitative insight. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:24:09 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Cui",
"C.",
""
],
[
"Cupertino-Malheiros",
"L.",
""
],
[
"Xiong",
"Z.",
""
],
[
"Martínez-Pañeda",
"E.",
""
]
] |
2506.02539 | Thong Q. Nguyen | Thong Q. Nguyen, Shubhang Desai, Yash Jain, Tanvir Aumi, Vishal Chowdhary | VerificAgent: Integrating Expert Knowledge and Fact-Checked Memory for Robust Domain-Specific Task Planning | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Continual memory augmentation allows computer-use agents (CUAs) to learn from past interactions and refine their task-solving strategies over time. However, unchecked memory accumulation can introduce spurious or hallucinated "learnings" that degrade agent performance, particularly in domain-specific workflows such as productivity software. We present a novel framework, VerificAgent, that effectively manages memory for CUAs through (1) an expert-curated seed of domain knowledge, (2) iterative, trajectory-based memory refinement during training, and (3) a post-hoc fact-checking pass by human experts to sanitize accumulated memory before deployment. On OSWorld productivity tasks, VerificAgent achieves a 111.1% relative improvement in success rate over baseline CUA without any additional fine-tuning. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:25:49 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Nguyen",
"Thong Q.",
""
],
[
"Desai",
"Shubhang",
""
],
[
"Jain",
"Yash",
""
],
[
"Aumi",
"Tanvir",
""
],
[
"Chowdhary",
"Vishal",
""
]
] |
2506.02540 | David Klompenhouwer | David Klompenhouwer, Stijn Velstra | Spin refinement of moduli spaces of residueless meromorphic differentials and the BKP hierarchy | 56 pages. Comments welcome. arXiv admin note: text overlap with arXiv:2110.01419 by other authors | null | null | null | math.AG math-ph math.MP nlin.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider strata of curves carrying a residueless meromorphic differential inducing a spin structure on the curve. The cohomology classes of the closures of these strata, weighted by the parity of the spin structures, form a partial cohomological field theory (CohFT) of infinite rank. After applying the DR hierarchy construction to this partial CohFT and reducing to differentials with two zeros and arbitrarily many poles, we show that the resulting system of evolutionary PDEs coincides with the BKP hierarchy up to a coordinate transformation. This is a spin refinement of an analogous result from arXiv:2110.01419. Our proof relies on a new result regarding the reconstruction of the BKP hierarchy from a limited amount of information in the Lax formalism. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:26:56 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Klompenhouwer",
"David",
""
],
[
"Velstra",
"Stijn",
""
]
] |
2506.02541 | Minsung Kim | Minsung Kim, Nakyeong Yang and Kyomin Jung | Rethinking Post-Unlearning Behavior of Large Vision-Language Models | 10 pages, 5 figures | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | Machine unlearning is used to mitigate the privacy risks of Large Vision-Language Models (LVLMs) arising from training on large-scale web data. However, existing unlearning methods often fail to carefully select substitute outputs for forget targets, resulting in Unlearning Aftermaths-undesirable behaviors such as degenerate, hallucinated, or excessively refused responses. We highlight that, especially for generative LVLMs, it is crucial to consider the quality and informativeness of post-unlearning responses rather than relying solely on naive suppression. To address this, we introduce a new unlearning task for LVLMs that requires models to provide privacy-preserving yet informative and visually grounded responses. We also propose PUBG, a novel unlearning method that explicitly guides post-unlearning behavior toward a desirable output distribution. Experiments show that, while existing methods suffer from Unlearning Aftermaths despite successfully preventing privacy violations, PUBG effectively mitigates these issues, generating visually grounded and informative responses without privacy leakage for forgotten targets. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:28:22 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kim",
"Minsung",
""
],
[
"Yang",
"Nakyeong",
""
],
[
"Jung",
"Kyomin",
""
]
] |
2506.02542 | Niklas Kormann | Niklas Kormann, Masoud Ramuz, Zeeshan Nisar, Nadine S. Schaadt, Hendrik Annuth, Benjamin Doerr, Friedrich Feuerhake, Thomas Lampert, Johannes F. Lutzeyer | HIEGNet: A Heterogenous Graph Neural Network Including the Immune Environment in Glomeruli Classification | Accepted for poster presentation at MIDL 2025 | null | null | null | cs.LG cs.AI cs.CV q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | Graph Neural Networks (GNNs) have recently been found to excel in histopathology. However, an important histopathological task, where GNNs have not been extensively explored, is the classification of glomeruli health as an important indicator in nephropathology. This task presents unique difficulties, particularly for the graph construction, i.e., the identification of nodes, edges, and informative features. In this work, we propose a pipeline composed of different traditional and machine learning-based computer vision techniques to identify nodes, edges, and their corresponding features to form a heterogeneous graph. We then proceed to propose a novel heterogeneous GNN architecture for glomeruli classification, called HIEGNet, that integrates both glomeruli and their surrounding immune cells. Hence, HIEGNet is able to consider the immune environment of each glomerulus in its classification. Our HIEGNet was trained and tested on a dataset of Whole Slide Images from kidney transplant patients. Experimental results demonstrate that HIEGNet outperforms several baseline models and generalises best between patients among all baseline models. Our implementation is publicly available at https://github.com/nklsKrmnn/HIEGNet.git. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:28:25 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kormann",
"Niklas",
""
],
[
"Ramuz",
"Masoud",
""
],
[
"Nisar",
"Zeeshan",
""
],
[
"Schaadt",
"Nadine S.",
""
],
[
"Annuth",
"Hendrik",
""
],
[
"Doerr",
"Benjamin",
""
],
[
"Feuerhake",
"Friedrich",
""
],
[
"Lampert",
"Thomas",
""
],
[
"Lutzeyer",
"Johannes F.",
""
]
] |
2506.02543 | Dinesh Chandra Maurya | Dinesh Chandra Maurya | Quintessence dark energy model in non-linear $f(Q)$ theory with bulk-viscosity | 17 pages, 11 figures | null | null | null | gr-qc | http://creativecommons.org/publicdomain/zero/1.0/ | In this study, we investigate a locally rotationally symmetric (LRS) Bianchi type-I cosmological model in non-linear form of $f(Q)$ gravity with observational constraints. We solved the modified Einstein's field equations with a viscous fluid source and got a hyperbolic solution. First, we apply MCMC analysis to the cosmic chronometer (CC) and Pantheon datasets to place observational constraints on model parameters. Using approximated model parameters, we study the cosmological parameters, such as the Hubble parameter $H$, the deceleration parameter $q$, and the equation of state (EoS) parameter $\omega_{v}$ with the skewness parameter $\delta_{v}$ for the viscous fluid. In addition, we examined Om diagnostics and statefinder analysis to categorize dark energy models. We estimated the age of the current universe as $t_{0}\approx13.8$ Gyrs. We obtained a quintessential and ever-accelerating model with bulk viscosity fluid. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:31:10 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Maurya",
"Dinesh Chandra",
""
]
] |
2506.02545 | Kemal Altwlkany | Kemal Altwlkany, Amar Kuric, Emanuel Lacic | On the Language and Gender Biases in PSTN, VoIP and Neural Audio Codecs | Submitted to INTERSPEECH2025 | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | In recent years, there has been a growing focus on fairness and inclusivity within speech technology, particularly in areas such as automatic speech recognition and speech sentiment analysis. When audio is transcoded prior to processing, as is the case in streaming or real-time applications, any inherent bias in the coding mechanism may result in disparities. This not only affects user experience but can also have broader societal implications by perpetuating stereotypes and exclusion. Thus, it is important that audio coding mechanisms are unbiased. In this work, we contribute towards the scarce research with respect to language and gender biases of audio codecs. By analyzing the speech quality of over 2 million multilingual audio files after transcoding through a representative subset of codecs (PSTN, VoIP and neural), our results indicate that PSTN codecs are strongly biased in terms of gender and that neural codecs introduce language biases. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:32:42 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Altwlkany",
"Kemal",
""
],
[
"Kuric",
"Amar",
""
],
[
"Lacic",
"Emanuel",
""
]
] |
2506.02546 | Pengfei He | Pengfei He, Zhenwei Dai, Xianfeng Tang, Yue Xing, Hui Liu, Jingying Zeng, Qiankun Peng, Shrivats Agrawal, Samarth Varshney, Suhang Wang, Jiliang Tang, Qi He | Attention Knows Whom to Trust: Attention-based Trust Management for LLM Multi-Agent Systems | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Large Language Model-based Multi-Agent Systems (LLM-MAS) have demonstrated strong capabilities in solving complex tasks but remain vulnerable when agents receive unreliable messages. This vulnerability stems from a fundamental gap: LLM agents treat all incoming messages equally without evaluating their trustworthiness. While some existing studies approach the trustworthiness, they focus on a single type of harmfulness rather than analyze it in a holistic approach from multiple trustworthiness perspectives. In this work, we propose Attention Trust Score (A-Trust), a lightweight, attention-based method for evaluating message trustworthiness. Inspired by human communication literature[1], through systematically analyzing attention behaviors across six orthogonal trust dimensions, we find that certain attention heads in the LLM specialize in detecting specific types of violations. Leveraging these insights, A-Trust directly infers trustworthiness from internal attention patterns without requiring external prompts or verifiers. Building upon A-Trust, we develop a principled and efficient trust management system (TMS) for LLM-MAS, enabling both message-level and agent-level trust assessment. Experiments across diverse multi-agent settings and tasks demonstrate that applying our TMS significantly enhances robustness against malicious inputs. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:32:57 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"He",
"Pengfei",
""
],
[
"Dai",
"Zhenwei",
""
],
[
"Tang",
"Xianfeng",
""
],
[
"Xing",
"Yue",
""
],
[
"Liu",
"Hui",
""
],
[
"Zeng",
"Jingying",
""
],
[
"Peng",
"Qiankun",
""
],
[
"Agrawal",
"Shrivats",
""
],
[
"Varshney",
"Samarth",
""
],
[
"Wang",
"Suhang",
""
],
[
"Tang",
"Jiliang",
""
],
[
"He",
"Qi",
""
]
] |
2506.02547 | Andreu Girbau Xalabarder | Andreu Girbau-Xalabarder, Jun Nagata, Shinichi Sumiyoshi | Probabilistic Online Event Downsampling | Accepted at CVPR 2025 Event-Vision workshop | null | null | null | cs.CV cs.ET | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Event cameras capture scene changes asynchronously on a per-pixel basis, enabling extremely high temporal resolution. However, this advantage comes at the cost of high bandwidth, memory, and computational demands. To address this, prior work has explored event downsampling, but most approaches rely on fixed heuristics or threshold-based strategies, limiting their adaptability. Instead, we propose a probabilistic framework, POLED, that models event importance through an event-importance probability density function (ePDF), which can be arbitrarily defined and adapted to different applications. Our approach operates in a purely online setting, estimating event importance on-the-fly from raw event streams, enabling scene-specific adaptation. Additionally, we introduce zero-shot event downsampling, where downsampled events must remain usable for models trained on the original event stream, without task-specific adaptation. We design a contour-preserving ePDF that prioritizes structurally important events and evaluate our method across four datasets and tasks--object classification, image interpolation, surface normal estimation, and object detection--demonstrating that intelligent sampling is crucial for maintaining performance under event-budget constraints. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:33:11 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Girbau-Xalabarder",
"Andreu",
""
],
[
"Nagata",
"Jun",
""
],
[
"Sumiyoshi",
"Shinichi",
""
]
] |
2506.02548 | Zhun Wang | Zhun Wang, Tianneng Shi, Jingxuan He, Matthew Cai, Jialin Zhang, Dawn Song | CyberGym: Evaluating AI Agents' Cybersecurity Capabilities with Real-World Vulnerabilities at Scale | null | null | null | null | cs.CR cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language model (LLM) agents are becoming increasingly skilled at handling cybersecurity tasks autonomously. Thoroughly assessing their cybersecurity capabilities is critical and urgent, given the high stakes in this domain. However, existing benchmarks fall short, often failing to capture real-world scenarios or being limited in scope. To address this gap, we introduce CyberGym, a large-scale and high-quality cybersecurity evaluation framework featuring 1,507 real-world vulnerabilities found and patched across 188 large software projects. While it includes tasks of various settings, CyberGym primarily focuses on the generation of proof-of-concept (PoC) tests for vulnerability reproduction, based on text descriptions and corresponding source repositories. Solving this task is particularly challenging, as it requires comprehensive reasoning across entire codebases to locate relevant code fragments and produce effective PoCs that accurately trigger the target vulnerability starting from the program's entry point. Our evaluation across 4 state-of-the-art agent frameworks and 9 LLMs reveals that even the best combination (OpenHands and Claude-3.7-Sonnet) achieves only a 11.9% reproduction success rate, mainly on simpler cases. Beyond reproducing historical vulnerabilities, we find that PoCs generated by LLM agents can reveal new vulnerabilities, identifying 15 zero-days affecting the latest versions of the software projects. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:35:14 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Wang",
"Zhun",
""
],
[
"Shi",
"Tianneng",
""
],
[
"He",
"Jingxuan",
""
],
[
"Cai",
"Matthew",
""
],
[
"Zhang",
"Jialin",
""
],
[
"Song",
"Dawn",
""
]
] |
2506.02549 | Akie Kowaguchi | Akie Kowaguchi, Katsuhiro Endo, Kentaro Nomura, Shuichi Kurabayashi, Paul E. Brumby, and Kenji Yasuoka | Navigating Phase Transitions with Path-Finding Algorithms: A Strategic Approach to Replica Exchange Monte Carlo | 7 pages | null | null | null | cond-mat.stat-mech physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | The replica exchange method is a powerful tool for overcoming slow relaxation in molecular simulations, but its efficiency depends strongly on the choice of the number and interval of replicas and their exchange probabilities. Here, we propose a new optimization scheme based on the Dijkstra algorithm that constructs an optimal exchange path by representing replicas and their exchange probabilities as a graph. Inspired by path-finding techniques widely used in computer science, including applications in game algorithms, our approach ensures that transitions follow a minimum entropy gradient path and effectively speeds up sampling even in systems exhibiting slow relaxation near critical points or phase transition regions. The method provides a systematic way to improve replica exchange efficiency and offers new insights into the control of relaxation dynamics, as demonstrated through applications to the solid-liquid phase transition of the Lennard-Jones bulk system. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:36:06 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kowaguchi",
"Akie",
""
],
[
"Endo",
"Katsuhiro",
""
],
[
"Nomura",
"Kentaro",
""
],
[
"Kurabayashi",
"Shuichi",
""
],
[
"Brumby",
"Paul E.",
""
],
[
"Yasuoka",
"Kenji",
""
]
] |
2506.02550 | Qiaohui Chu | Qiaohui Chu, Haoyu Zhang, Yisen Feng, Meng Liu, Weili Guan, Yaowei Wang, Liqiang Nie | Technical Report for Ego4D Long-Term Action Anticipation Challenge 2025 | The champion solution for the Ego4D Long-Term Action Anticipation Challenge at the CVPR EgoVis Workshop 2025 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this report, we present a novel three-stage framework developed for the Ego4D Long-Term Action Anticipation (LTA) task. Inspired by recent advances in foundation models, our method consists of three stages: feature extraction, action recognition, and long-term action anticipation. First, visual features are extracted using a high-performance visual encoder. The features are then fed into a Transformer to predict verbs and nouns, with a verb-noun co-occurrence matrix incorporated to enhance recognition accuracy. Finally, the predicted verb-noun pairs are formatted as textual prompts and input into a fine-tuned large language model (LLM) to anticipate future action sequences. Our framework achieves first place in this challenge at CVPR 2025, establishing a new state-of-the-art in long-term action prediction. Our code will be released at https://github.com/CorrineQiu/Ego4D-LTA-Challenge-2025. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:36:52 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Chu",
"Qiaohui",
""
],
[
"Zhang",
"Haoyu",
""
],
[
"Feng",
"Yisen",
""
],
[
"Liu",
"Meng",
""
],
[
"Guan",
"Weili",
""
],
[
"Wang",
"Yaowei",
""
],
[
"Nie",
"Liqiang",
""
]
] |
2506.02552 | Fulvia De Fazio | Pietro Colangelo, Fulvia De Fazio and Davide Milillo | Correlating lepton flavour violating $b \to s$ and leptonic decay modes in a minimal abelian extension of the Standard Model | 38 pages, 20 figures | null | null | BARI-TH/25-772 | hep-ph hep-ex hep-lat | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider an abelian extension of the Standard Model (SM) comprising a new gauge group $U(1)^\prime$, with the neutral gauge boson $Z^\prime$ having flavour violating couplings to quarks and leptons. The fermion content is the same as in SM except for the addition of three right-handed neutrinos. The model describes the couplings of $Z^\prime$ to fermions in terms of three rational parameters $\epsilon_{1,2,3}$ that sum to zero imposing the cancellation of the gauge anomalies. Each $\epsilon_i$ is common to all fermions in a generation, a feature producing correlations among quark and lepton observables. We focus on $b \to s \ell_1^- \ell_2^+$ transitions for the lepton flavour conserving $\ell_1=\ell_2$ and lepton flavour violating case $\ell_1 \neq \ell_2$. Small deviations with respect to the SM predictions are found in the first case, which reflects a feature of the model where quark and lepton sectors prevent each other to manifest large discrepancies with respect to SM. We investigate the correlations with the flavour violating leptonic decays $\mu^- \to e^- \gamma$, $\tau^- \to \mu^- \mu^+ \mu^-$. The experimental upper bound on the branching ratio of $\mu^- \to e^- \gamma$ constrains the range for the lepton flavour violating $B_{d,s}$ decays, which however are predicted to be within the reach of the planned new facilities. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:40:26 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Colangelo",
"Pietro",
""
],
[
"De Fazio",
"Fulvia",
""
],
[
"Milillo",
"Davide",
""
]
] |
2506.02553 | Shenghua He | Shenghua He and Tian Xia and Xuan Zhou and Hui Wei | Response-Level Rewards Are All You Need for Online Reinforcement Learning in LLMs: A Mathematical Perspective | null | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | We study a common challenge in reinforcement learning for large language models (LLMs): the Zero-Reward Assumption, where non-terminal actions (i.e., intermediate token generations) receive zero task-specific immediate reward, while only the final token receives a reward for the entire response. This assumption arises frequently in practice, as precise token-level rewards are often difficult or infeasible to obtain in LLM applications. In this work, we provide a unifying theoretical perspective. We introduce the Trajectory Policy Gradient Theorem, which shows that the policy gradient based on true, unknown token-level rewards can be unbiasedly estimated using only a response-level reward model, regardless of whether the Zero-Reward Assumption holds or not, for algorithms in the REINFORCE and Actor-Critic families. This result reveals that widely used methods such as PPO, GRPO, ReMax, and RLOO inherently possess the capacity to model token-level reward signals, offering a theoretical justification for response-level reward approaches. Our findings pave the way for more practical, efficient LLM fine-tuning, allowing developers to treat training algorithms as black boxes and focus on improving the response-level reward model with auxiliary sub-models. We also offer a detailed analysis of popular RL and non-RL methods, comparing their theoretical foundations and practical advantages across common LLM tasks. Finally, we propose a new algorithm: Token-Reinforced Policy Optimization (TRePO), a theoretically grounded method that is simpler than PPO, matches GRPO in memory efficiency, and holds promise for broad applicability. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:44:31 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"He",
"Shenghua",
""
],
[
"Xia",
"Tian",
""
],
[
"Zhou",
"Xuan",
""
],
[
"Wei",
"Hui",
""
]
] |
2506.02554 | Timo Osterburg | Timo Osterburg, Franz Albers, Christopher Diehl, Rajesh Pushparaj and Torsten Bertram | HiLO: High-Level Object Fusion for Autonomous Driving using Transformers | 6 pages, accepted at IEEE Intelligent Vehicles Symposium (IV) 2025 | null | null | null | cs.RO cs.AI cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The fusion of sensor data is essential for a robust perception of the environment in autonomous driving. Learning-based fusion approaches mainly use feature-level fusion to achieve high performance, but their complexity and hardware requirements limit their applicability in near-production vehicles. High-level fusion methods offer robustness with lower computational requirements. Traditional methods, such as the Kalman filter, dominate this area. This paper modifies the Adapted Kalman Filter (AKF) and proposes a novel transformer-based high-level object fusion method called HiLO. Experimental results demonstrate improvements of $25.9$ percentage points in $\textrm{F}_1$ score and $6.1$ percentage points in mean IoU. Evaluation on a new large-scale real-world dataset demonstrates the effectiveness of the proposed approaches. Their generalizability is further validated by cross-domain evaluation between urban and highway scenarios. Code, data, and models are available at https://github.com/rst-tu-dortmund/HiLO . | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:44:35 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Osterburg",
"Timo",
""
],
[
"Albers",
"Franz",
""
],
[
"Diehl",
"Christopher",
""
],
[
"Pushparaj",
"Rajesh",
""
],
[
"Bertram",
"Torsten",
""
]
] |
2506.02555 | Zhitao Zeng | Zhitao Zeng and Zhu Zhuo and Xiaojun Jia and Erli Zhang and Junde Wu and Jiaan Zhang and Yuxuan Wang and Chang Han Low and Jian Jiang and Zilong Zheng and Xiaochun Cao and Yutong Ban and Qi Dou and Yang Liu and Yueming Jin | SurgVLM: A Large Vision-Language Model and Systematic Evaluation Benchmark for Surgical Intelligence | 29 pages, 5 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Foundation models have achieved transformative success across biomedical domains by enabling holistic understanding of multimodal data. However, their application in surgery remains underexplored. Surgical intelligence presents unique challenges - requiring surgical visual perception, temporal analysis, and reasoning. Existing general-purpose vision-language models fail to address these needs due to insufficient domain-specific supervision and the lack of a large-scale high-quality surgical database. To bridge this gap, we propose SurgVLM, one of the first large vision-language foundation models for surgical intelligence, where this single universal model can tackle versatile surgical tasks. To enable this, we construct a large-scale multimodal surgical database, SurgVLM-DB, comprising over 1.81 million frames with 7.79 million conversations, spanning more than 16 surgical types and 18 anatomical structures. We unify and reorganize 23 public datasets across 10 surgical tasks, followed by standardizing labels and doing hierarchical vision-language alignment to facilitate comprehensive coverage of gradually finer-grained surgical tasks, from visual perception, temporal analysis, to high-level reasoning. Building upon this comprehensive dataset, we propose SurgVLM, which is built upon Qwen2.5-VL, and undergoes instruction tuning to 10+ surgical tasks. We further construct a surgical multimodal benchmark, SurgVLM-Bench, for method evaluation. SurgVLM-Bench consists of 6 popular and widely-used datasets in surgical domain, covering several crucial downstream tasks. Based on SurgVLM-Bench, we evaluate the performance of our SurgVLM (3 SurgVLM variants: SurgVLM-7B, SurgVLM-32B, and SurgVLM-72B), and conduct comprehensive comparisons with 14 mainstream commercial VLMs (e.g., GPT-4o, Gemini 2.0 Flash, Qwen2.5-Max). | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:44:41 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zeng",
"Zhitao",
""
],
[
"Zhuo",
"Zhu",
""
],
[
"Jia",
"Xiaojun",
""
],
[
"Zhang",
"Erli",
""
],
[
"Wu",
"Junde",
""
],
[
"Zhang",
"Jiaan",
""
],
[
"Wang",
"Yuxuan",
""
],
[
"Low",
"Chang Han",
""
],
[
"Jiang",
"Jian",
""
],
[
"Zheng",
"Zilong",
""
],
[
"Cao",
"Xiaochun",
""
],
[
"Ban",
"Yutong",
""
],
[
"Dou",
"Qi",
""
],
[
"Liu",
"Yang",
""
],
[
"Jin",
"Yueming",
""
]
] |
2506.02556 | Nicky Zimmerman | Ayush Agrawal, Joel Loo, Nicky Zimmerman, David Hsu | Sign Language: Towards Sign Understanding for Robot Autonomy | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Signage is an ubiquitous element of human environments, playing a critical role in both scene understanding and navigation. For autonomous systems to fully interpret human environments, effectively parsing and understanding signs is essential. We introduce the task of navigational sign understanding, aimed at extracting navigational cues from signs that convey symbolic spatial information about the scene. Specifically, we focus on signs capturing directional cues that point toward distant locations and locational cues that identify specific places. To benchmark performance on this task, we curate a comprehensive test set, propose appropriate evaluation metrics, and establish a baseline approach. Our test set consists of over 160 images, capturing signs with varying complexity and design across a wide range of public spaces, such as hospitals, shopping malls, and transportation hubs. Our baseline approach harnesses Vision-Language Models (VLMs) to parse navigational signs under these high degrees of variability. Experiments show that VLMs offer promising performance on this task, potentially motivating downstream applications in robotics. The code and dataset are available on Github. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:44:42 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Agrawal",
"Ayush",
""
],
[
"Loo",
"Joel",
""
],
[
"Zimmerman",
"Nicky",
""
],
[
"Hsu",
"David",
""
]
] |
2506.02557 | Shizhan Gong | Shizhan Gong, Yankai Jiang, Qi Dou, Farzan Farnia | Kernel-based Unsupervised Embedding Alignment for Enhanced Visual Representation in Vision-language Models | ICML 2025 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Vision-language models, such as CLIP, have achieved significant success in aligning visual and textual representations, becoming essential components of many multi-modal large language models (MLLMs) like LLaVA and OpenFlamingo. However, numerous studies have identified CLIP's limited fine-grained perception as a critical drawback, leading to substantial failures in downstream MLLMs. In contrast, vision-centric foundation models like DINOv2 demonstrate remarkable capabilities in capturing fine details from images. In this work, we propose a novel kernel-based method to align CLIP's visual representation with that of DINOv2, ensuring that the resulting embeddings maintain compatibility with text embeddings while enhancing perceptual capabilities. Our alignment objective is designed for efficient stochastic optimization. Following this image-only alignment fine-tuning, the visual encoder retains compatibility with the frozen text encoder and exhibits significant improvements in zero-shot object recognition, fine-grained spatial reasoning, and localization. By integrating the aligned visual encoder, downstream MLLMs also demonstrate enhanced performance. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:44:43 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Gong",
"Shizhan",
""
],
[
"Jiang",
"Yankai",
""
],
[
"Dou",
"Qi",
""
],
[
"Farnia",
"Farzan",
""
]
] |
2506.02558 | Pramod Sharma | Pramod Sharma and Ambresh Shivaji | Role of angular observables in probing non-standard $HZZ$ couplings at an electron-proton collider | 18 pages, 6 figures | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this study, we explore various lab-frame angular observables at Large Hadron-electron Collider (LHeC) to test their sensitivity on the most general structure of the Higgs ($H$) to neutral weak boson ($Z$) coupling ($HZZ$) via the process $e^- p \rightarrow e^- H j$ at the center of mass energy $\sqrt{s}~\approx$ 1.3 TeV. The most general Lorentz structure of the $HZZ$ coupling beyond the standard model is composed of two CP-even ($\lambda_{1Z}$ and $\lambda_{2Z}$) and one CP-odd (${\tilde \lambda_Z}$) components. In a previous study, we looked at the absolute value of azimuthal correlation ($|\Delta \phi|$) between the final state electron and the jet to derive constraints on the non-standard $HZZ$ couplings. This choice of observable is motivated by its potential to discriminate between CP-odd and CP-even couplings in the $e^- p \rightarrow \nu_e H j$ process. Since the process $e^- p \rightarrow e^- H j$ has an electron in the final state, one can construct more angular observables. In addition to $|\Delta \phi|$, we identify new angular observables: the sign-sensitive azimuthal correlation ($\Delta \phi$) and the polar angle of the final state electron ($\theta$), suitable for constraining non-standard $HZZ$ coupling. Our analysis shows that these new angles improve the constraints on $\lambda_{2Z}$ and ${\tilde \lambda}_Z$ in the range of 48-67\%. We also study the potential of the asymmetry in $\Delta \phi$ to constrain the parameters corresponding to the CP-odd coupling. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:45:12 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Sharma",
"Pramod",
""
],
[
"Shivaji",
"Ambresh",
""
]
] |
2506.02559 | Fatih Kansoy | Fatih Kansoy, Joel Mundy | Central Bank Communication with Public: Bank of England and Twitter (X) | 35 pages, 11 figures | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | Central banks increasingly use social media to communicate beyond financial markets, yet evidence on public engagement effectiveness remains limited. Despite 113 central banks joining Twitter between 2008 and 2018, we lack understanding of what drives audience interaction with their content. To examine engagement determinants, we analyzed 3.13 million tweets mentioning the Bank of England from 2007 to 2022, including 9,810 official posts. We investigate posting patterns, measure engagement elasticity, and identify content characteristics predicting higher interaction. The Bank's posting schedule misaligns with peak audience engagement times, with evening hours generating the highest interaction despite minimal posting. Cultural content, such as the Alan Turing 50 pound note, achieved 1,300 times higher engagement than routine policy communications. Engagement elasticity averaged 1.095 with substantial volatility during events like Brexit, contrasting with the Federal Reserve's stability. Media content dramatically increased engagement: videos by 1,700 percent, photos by 126 percent, while monetary policy announcements and readability significantly enhanced all metrics. Content quality and timing matter more than posting frequency for effective central bank communication. These findings suggest central banks should prioritize accessible, media-rich content during high-attention periods rather than increasing volume, with implications for digital communication strategies in fulfilling public transparency mandates. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:46:24 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kansoy",
"Fatih",
""
],
[
"Mundy",
"Joel",
""
]
] |
2506.02560 | Zixiang Li | Zixiang Li, Haoyu Wang, Wei Wang, Chuangchuang Tan, Yunchao Wei, Yao Zhao | DCI: Dual-Conditional Inversion for Boosting Diffusion-Based Image Editing | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Diffusion models have achieved remarkable success in image generation and editing tasks. Inversion within these models aims to recover the latent noise representation for a real or generated image, enabling reconstruction, editing, and other downstream tasks. However, to date, most inversion approaches suffer from an intrinsic trade-off between reconstruction accuracy and editing flexibility. This limitation arises from the difficulty of maintaining both semantic alignment and structural consistency during the inversion process. In this work, we introduce Dual-Conditional Inversion (DCI), a novel framework that jointly conditions on the source prompt and reference image to guide the inversion process. Specifically, DCI formulates the inversion process as a dual-condition fixed-point optimization problem, minimizing both the latent noise gap and the reconstruction error under the joint guidance. This design anchors the inversion trajectory in both semantic and visual space, leading to more accurate and editable latent representations. Our novel setup brings new understanding to the inversion process. Extensive experiments demonstrate that DCI achieves state-of-the-art performance across multiple editing tasks, significantly improving both reconstruction quality and editing precision. Furthermore, we also demonstrate that our method achieves strong results in reconstruction tasks, implying a degree of robustness and generalizability approaching the ultimate goal of the inversion process. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:46:44 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Li",
"Zixiang",
""
],
[
"Wang",
"Haoyu",
""
],
[
"Wang",
"Wei",
""
],
[
"Tan",
"Chuangchuang",
""
],
[
"Wei",
"Yunchao",
""
],
[
"Zhao",
"Yao",
""
]
] |
2506.02561 | Yiran Zhao | Yirao Zhao, Guizhen Chen, Kenji Kawaguchi, Lidong Bing, Wenxuan Zhang | Pruning General Large Language Models into Customized Expert Models | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have revolutionized natural language processing, yet their substantial model sizes often require substantial computational resources. To preserve computing resources and accelerate inference speed, it is crucial to prune redundant parameters, especially for experienced users who often need compact expert models tailored to specific downstream scenarios. However, most existing pruning methods focus on preserving the model's general capabilities, often requiring extensive post-training or suffering from degraded performance due to coarse-grained pruning. In this work, we design a $\underline{Cus}$tom $\underline{Prun}$ing method ($\texttt{Cus-Prun}$) to prune a large general model into a smaller lightweight expert model, which is positioned along the "language", "domain" and "task" dimensions. By identifying and pruning irrelevant neurons of each dimension, $\texttt{Cus-Prun}$ creates expert models without any post-training. Our experiments demonstrate that $\texttt{Cus-Prun}$ consistently outperforms other methods, achieving minimal loss in both expert and general capabilities across various models from different model families and sizes. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:47:30 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zhao",
"Yirao",
""
],
[
"Chen",
"Guizhen",
""
],
[
"Kawaguchi",
"Kenji",
""
],
[
"Bing",
"Lidong",
""
],
[
"Zhang",
"Wenxuan",
""
]
] |
2506.02562 | Jie Li | Xuan Zuo, Zi-Xu Lu, Shi-Yao Zhu, Jie Li | Polaromechanical cooling of multiple mechanical modes | null | null | null | null | quant-ph cond-mat.mes-hall physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The strong coupling between light and matter gives rise to polaritons. Further coupling polaritons to phonons leads to the formation of hybrid polaromechanical systems. Recent experiments have achieved the strong coupling between polaritons and phonons in two different configurations, namely, the magnon-photon-phonon and exciton-photon-phonon systems, which makes it possible to prepare and control mechanical states via manipulating the polaritons. Here, we present a polaromechanical cooling theory and show how multiple mechanical modes can be simultaneously cooled by using polaritons with a single drive field. This is achieved by exploiting the great tunability of the polariton frequencies. Specifically, $N$ polaritons can be tuned to resonantly enhance the anti-Stokes scatterings associated with $N$ mechanical modes, leading to the cooling of all the mechanical modes at once. We explicitly show how such polaromechanical cooling can simultaneously cool two mechanical modes, and then discuss how to apply it to cool multiple mechanical modes and to various polaromechanical systems, such as magnon-, exciton-, and plasmon-polaromechanical systems. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:47:47 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Zuo",
"Xuan",
""
],
[
"Lu",
"Zi-Xu",
""
],
[
"Zhu",
"Shi-Yao",
""
],
[
"Li",
"Jie",
""
]
] |
2506.02563 | Roie Reshef | Roie Reshef and Kfir Yehuda Levy | Privacy-Preserving Federated Convex Optimization: Balancing Partial-Participation and Efficiency via Noise Cancellation | arXiv admin note: text overlap with arXiv:2407.12396 | ICML 2025 | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper tackles the challenge of achieving Differential Privacy (DP) in Federated Learning (FL) under partial-participation, where only a subset of the machines participate in each time-step. While previous work achieved optimal performance in full-participation settings, these methods struggled to extend to partial-participation scenarios. Our approach fills this gap by introducing a novel noise-cancellation mechanism that preserves privacy without sacrificing convergence rates or computational efficiency. We analyze our method within the Stochastic Convex Optimization (SCO) framework and show that it delivers optimal performance for both homogeneous and heterogeneous data distributions. This work expands the applicability of DP in FL, offering an efficient and practical solution for privacy-preserving learning in distributed systems with partial participation. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:48:35 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Reshef",
"Roie",
""
],
[
"Levy",
"Kfir Yehuda",
""
]
] |
2506.02564 | David \v{S}i\v{s}ka | Deven Sethi, David \v{S}i\v{s}ka | Mirror descent for constrained stochastic control problems | null | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mirror descent is a well established tool for solving convex optimization problems with convex constraints. This article introduces continuous-time mirror descent dynamics for approximating optimal Markov controls for stochastic control problems with the action space being bounded and convex. We show that if the Hamiltonian is uniformly convex in its action variable then mirror descent converges linearly while if it is uniformly strongly convex relative to an appropriate Bregman divergence, then the mirror flow converges exponentially. The two fundamental difficulties that must be overcome to prove such results are: first, the inherent lack of convexity of the map from Markov controls to the corresponding value function. Second, maintaining sufficient regularity of the value function and the Markov controls along the mirror descent updates. The first issue is handled using the performance difference lemma, while the second using careful Sobolev space estimates for the solutions of the associated linear PDEs. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:49:17 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Sethi",
"Deven",
""
],
[
"Šiška",
"David",
""
]
] |
2506.02565 | Zhuoxuan Jiang | Zhuoxuan Jiang, Tianyang Zhang, Peiyan Peng, Jing Chen, Yinong Xun, Haotian Zhang, Lichi Li, Yong Li, Shaohua Zhang | Towards Generating Controllable and Solvable Geometry Problem by Leveraging Symbolic Deduction Engine | To Appear in ACL'25 | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generating high-quality geometry problems is both an important and challenging task in education. Compared to math word problems, geometry problems further emphasize multi-modal formats and the translation between informal and formal languages. In this paper, we introduce a novel task for geometry problem generation and propose a new pipeline method: the Symbolic Deduction Engine-based Geometry Problem Generation framework (SDE-GPG). The framework leverages a symbolic deduction engine and contains four main steps: (1) searching a predefined mapping table from knowledge points to extended definitions, (2) sampling extended definitions and performing symbolic deduction, (3) filtering out unqualified problems, and (4) generating textual problems and diagrams. Specifically, our method supports to avoid inherent biases in translating natural language into formal language by designing the mapping table, and guarantees to control the generated problems in terms of knowledge points and difficulties by an elaborate checking function. With obtained formal problems, they are translated to natural language and the accompanying diagrams are automatically drew by rule-based methods. We conduct experiments using real-world combinations of knowledge points from two public datasets. The results demonstrate that the SDE-GPG can effectively generate readable, solvable and controllable geometry problems. | [
{
"version": "v1",
"created": "Tue, 03 Jun 2025 07:49:38 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Jiang",
"Zhuoxuan",
""
],
[
"Zhang",
"Tianyang",
""
],
[
"Peng",
"Peiyan",
""
],
[
"Chen",
"Jing",
""
],
[
"Xun",
"Yinong",
""
],
[
"Zhang",
"Haotian",
""
],
[
"Li",
"Lichi",
""
],
[
"Li",
"Yong",
""
],
[
"Zhang",
"Shaohua",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.