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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.10334 | Deliang Wang | Deliang Wang and Chao Yang and Gaowei Chen | Using Vision Language Models to Detect Students' Academic Emotion through Facial Expressions | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Students' academic emotions significantly influence their social behavior and learning performance. Traditional approaches to automatically and accurately analyze these emotions have predominantly relied on supervised machine learning algorithms. However, these models often struggle to generalize across different contexts, necessitating repeated cycles of data collection, annotation, and training. The emergence of Vision-Language Models (VLMs) offers a promising alternative, enabling generalization across visual recognition tasks through zero-shot prompting without requiring fine-tuning. This study investigates the potential of VLMs to analyze students' academic emotions via facial expressions in an online learning environment. We employed two VLMs, Llama-3.2-11B-Vision-Instruct and Qwen2.5-VL-7B-Instruct, to analyze 5,000 images depicting confused, distracted, happy, neutral, and tired expressions using zero-shot prompting. Preliminary results indicate that both models demonstrate moderate performance in academic facial expression recognition, with Qwen2.5-VL-7B-Instruct outperforming Llama-3.2-11B-Vision-Instruct. Notably, both models excel in identifying students' happy emotions but fail to detect distracted behavior. Additionally, Qwen2.5-VL-7B-Instruct exhibits relatively high performance in recognizing students' confused expressions, highlighting its potential for practical applications in identifying content that causes student confusion. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:01:26 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wang",
"Deliang",
""
],
[
"Yang",
"Chao",
""
],
[
"Chen",
"Gaowei",
""
]
] |
2506.10335 | Lintao Xiang | Lintao Xiang, Hongpei Zheng, Yating Huang, Qijun Yang, Hujun Yin | PointGS: Point Attention-Aware Sparse View Synthesis with Gaussian Splatting | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D Gaussian splatting (3DGS) is an innovative rendering technique that surpasses the neural radiance field (NeRF) in both rendering speed and visual quality by leveraging an explicit 3D scene representation. Existing 3DGS approaches require a large number of calibrated views to generate a consistent and complete scene representation. When input views are limited, 3DGS tends to overfit the training views, leading to noticeable degradation in rendering quality. To address this limitation, we propose a Point-wise Feature-Aware Gaussian Splatting framework that enables real-time, high-quality rendering from sparse training views. Specifically, we first employ the latest stereo foundation model to estimate accurate camera poses and reconstruct a dense point cloud for Gaussian initialization. We then encode the colour attributes of each 3D Gaussian by sampling and aggregating multiscale 2D appearance features from sparse inputs. To enhance point-wise appearance representation, we design a point interaction network based on a self-attention mechanism, allowing each Gaussian point to interact with its nearest neighbors. These enriched features are subsequently decoded into Gaussian parameters through two lightweight multi-layer perceptrons (MLPs) for final rendering. Extensive experiments on diverse benchmarks demonstrate that our method significantly outperforms NeRF-based approaches and achieves competitive performance under few-shot settings compared to the state-of-the-art 3DGS methods. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:07:07 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Xiang",
"Lintao",
""
],
[
"Zheng",
"Hongpei",
""
],
[
"Huang",
"Yating",
""
],
[
"Yang",
"Qijun",
""
],
[
"Yin",
"Hujun",
""
]
] |
2506.10336 | Maoqiang Dong | M. Q. Dong, Zhi-Xin Guo, and Xin-Gao Gong | Field-free perpendicular magnetization switching by altermagnet with collinear spin current | null | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The generation of collinear spin current (CSC), where both the propagation direction and spin-polarized direction aligned perpendicularly to the applied charge current, is crucial for efficiently manipulating systems with perpendicular magnetic anisotropy used in high-density magnetic recording. However, the efficient generation of CSC remains a challenge. In this work, based on the symmetry analysis, we propose that CSC can be effectively generated using altermagnets when the charge current is aligned along specific directions, due to spin-dependent symmetry breaking. This proposal is supported by density functional theory (DFT) and Boltzmann transport equation (BTE) calculations on a series of altermagnetic materials, including RuO2, Mn5Si3, KRu4O8 and CuF2, where unusually large CSC is produced by the charge current along certain orientations. Furthermore, we introduce a physical quantity, the spin-splitting angle, to quantify the efficiency of CSC generated by the charge current. We find that the spin-splitting angle ranges from 0.24 to 0.57 in these altermagnets, which is significantly larger than the spin-Hall angle typically observed in the anomalous spin-Hall effect, where the spin-Hall angle is generally less than 0.1. Our findings provide an effective method for manipulating spin currents, which is advantageous for the exploration of altermagnetic spintronic devices with field-free perpendicular magnetization switching. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:19:21 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Dong",
"M. Q.",
""
],
[
"Guo",
"Zhi-Xin",
""
],
[
"Gong",
"Xin-Gao",
""
]
] |
2506.10337 | Zhanwei Zhang | Zhanwei Zhang, Kaiyuan Liu, Junjie Liu, Wenxiao Wang, Binbin Lin, Liang Xie, Chen Shen, Deng Cai | GeoCAD: Local Geometry-Controllable CAD Generation | 18 pages, 12 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Local geometry-controllable computer-aided design (CAD) generation aims to modify local parts of CAD models automatically, enhancing design efficiency. It also ensures that the shapes of newly generated local parts follow user-specific geometric instructions (e.g., an isosceles right triangle or a rectangle with one corner cut off). However, existing methods encounter challenges in achieving this goal. Specifically, they either lack the ability to follow textual instructions or are unable to focus on the local parts. To address this limitation, we introduce GeoCAD, a user-friendly and local geometry-controllable CAD generation method. Specifically, we first propose a complementary captioning strategy to generate geometric instructions for local parts. This strategy involves vertex-based and VLLM-based captioning for systematically annotating simple and complex parts, respectively. In this way, we caption $\sim$221k different local parts in total. In the training stage, given a CAD model, we randomly mask a local part. Then, using its geometric instruction and the remaining parts as input, we prompt large language models (LLMs) to predict the masked part. During inference, users can specify any local part for modification while adhering to a variety of predefined geometric instructions. Extensive experiments demonstrate the effectiveness of GeoCAD in generation quality, validity and text-to-CAD consistency. Code will be available at https://github.com/Zhanwei-Z/GeoCAD. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:23:31 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Zhanwei",
""
],
[
"Liu",
"Kaiyuan",
""
],
[
"Liu",
"Junjie",
""
],
[
"Wang",
"Wenxiao",
""
],
[
"Lin",
"Binbin",
""
],
[
"Xie",
"Liang",
""
],
[
"Shen",
"Chen",
""
],
[
"Cai",
"Deng",
""
]
] |
2506.10338 | Kwangsu Lee | Kwangsu Lee | Adaptive Chosen-Ciphertext Security of Distributed Broadcast Encryption | arXiv admin note: text overlap with arXiv:2505.17527 | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Distributed broadcast encryption (DBE) is a specific kind of broadcast encryption (BE) where users independently generate their own public and private keys, and a sender can efficiently create a ciphertext for a subset of users by using the public keys of the subset users. Previously proposed DBE schemes have been proven in the adaptive chosen-plaintext attack (CPA) security model and have the disadvantage of requiring linear number of pairing operations when verifying the public key of a user. In this paper, we propose an efficient DBE scheme in bilinear groups and prove adaptive chosen-ciphertext attack (CCA) security for the first time. To do this, we first propose a semi-static CCA secure DBE scheme and prove the security under the $q$-Type assumption. Then, by modifying the generic transformation of Gentry and Waters that converts a semi-static CPA secure DBE scheme into an adaptive CPA secure DBE scheme to be applied to CCA secure DBE schemes, we propose an adaptive CCA secure DBE scheme and prove its adaptive CCA security. Our proposed DBE scheme is efficient because it requires constant size ciphertexts, constant size private keys, and linear size public keys, and the public key verification requires only a constant number of pairing operations and efficient group membership checks. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:27:14 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lee",
"Kwangsu",
""
]
] |
2506.10339 | Danny Segev | Noga Alon and Danny Segev | New Approximation Guarantees for The Inventory Staggering Problem | null | null | null | null | cs.DS | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Since its inception in the mid-60s, the inventory staggering problem has been explored and exploited in a wide range of application domains, such as production planning, stock control systems, warehousing, and aerospace/defense logistics. However, even with a rich history of academic focus, we are still very much in the dark when it comes to cornerstone computational questions around inventory staggering and to related structural characterizations, with our methodological toolbox being severely under-stocked.
The central contribution of this paper consists in devising a host of algorithmic techniques and analytical ideas -- some being entirely novel and some leveraging well-studied concepts in combinatorics and number theory -- for surpassing essentially all known approximation guarantees for the inventory staggering problem. In particular, our work demonstrates that numerous structural properties open the door for designing polynomial-time approximation schemes, including polynomially-bounded cycle lengths, constantly-many distinct time intervals, so-called nested instances, and pairwise coprime settings. These findings offer substantial improvements over currently available constant-factor approximations and resolve outstanding open questions in their respective contexts. In parallel, we develop new theory around a number of yet-uncharted questions, related to the sampling complexity of peak inventory estimation as well as to the plausibility of groupwise synchronization. Interestingly, we establish the global nature of inventory staggering, proving that there are $n$-item instances where, for every subset of roughly $\sqrt{n}$ items, no policy improves on the worst-possible one by a factor greater than $1+\epsilon$, whereas for the entire instance, there exists a policy that outperforms the worst-possible one by a factor of nearly $2$, which is optimal. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:29:43 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Alon",
"Noga",
""
],
[
"Segev",
"Danny",
""
]
] |
2506.10340 | Eric Gao | Eric Gao | Seeding an Uncertain Technology | 12 pages | null | null | null | econ.TH | http://creativecommons.org/licenses/by-nc-sa/4.0/ | I study how a startup with uncertainty over product quality and no knowledge of the underlying diffusion network optimally chooses initial seeds. To ensure widespread adoption when the product is good while minimizing negative perceptions when it is bad, the optimal number of initial seeds should grow logarithmically with network size. When there are agents of different types that govern their connectivity, it is asymptotically optimal to seed agents of a single type: the type that minimizes the marginal cost per probability of making the product go viral. These results rationalize startup behavior in practice. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:33:30 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Gao",
"Eric",
""
]
] |
2506.10341 | Wanqiao Xu | Wanqiao Xu, Allen Nie, Ruijie Zheng, Aditya Modi, Adith Swaminathan, Ching-An Cheng | Provably Learning from Language Feedback | null | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | Interactively learning from observation and language feedback is an increasingly studied area driven by the emergence of large language model (LLM) agents. While impressive empirical demonstrations have been shown, so far a principled framing of these decision problems remains lacking. In this paper, we formalize the Learning from Language Feedback (LLF) problem, assert sufficient assumptions to enable learning despite latent rewards, and introduce $\textit{transfer eluder dimension}$ as a complexity measure to characterize the hardness of LLF problems. We show that transfer eluder dimension captures the intuition that information in the feedback changes the learning complexity of the LLF problem. We demonstrate cases where learning from rich language feedback can be exponentially faster than learning from reward. We develop a no-regret algorithm, called $\texttt{HELiX}$, that provably solves LLF problems through sequential interactions, with performance guarantees that scale with the transfer eluder dimension of the problem. Across several empirical domains, we show that $\texttt{HELiX}$ performs well even when repeatedly prompting LLMs does not work reliably. Our contributions mark a first step towards designing principled interactive learning algorithms from generic language feedback. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:35:02 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Xu",
"Wanqiao",
""
],
[
"Nie",
"Allen",
""
],
[
"Zheng",
"Ruijie",
""
],
[
"Modi",
"Aditya",
""
],
[
"Swaminathan",
"Adith",
""
],
[
"Cheng",
"Ching-An",
""
]
] |
2506.10342 | Jun Yin | Jun Yin, Jing Zhong, Peilin Li, Pengyu Zeng, Miao Zhang, Ran Luo, Shuai Lu | UrbanSense:AFramework for Quantitative Analysis of Urban Streetscapes leveraging Vision Large Language Models | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Urban cultures and architectural styles vary significantly across cities due to geographical, chronological, historical, and socio-political factors. Understanding these differences is essential for anticipating how cities may evolve in the future. As representative cases of historical continuity and modern innovation in China, Beijing and Shenzhen offer valuable perspectives for exploring the transformation of urban streetscapes. However, conventional approaches to urban cultural studies often rely on expert interpretation and historical documentation, which are difficult to standardize across different contexts. To address this, we propose a multimodal research framework based on vision-language models, enabling automated and scalable analysis of urban streetscape style differences. This approach enhances the objectivity and data-driven nature of urban form research. The contributions of this study are as follows: First, we construct UrbanDiffBench, a curated dataset of urban streetscapes containing architectural images from different periods and regions. Second, we develop UrbanSense, the first vision-language-model-based framework for urban streetscape analysis, enabling the quantitative generation and comparison of urban style representations. Third, experimental results show that Over 80% of generated descriptions pass the t-test (p less than 0.05). High Phi scores (0.912 for cities, 0.833 for periods) from subjective evaluations confirm the method's ability to capture subtle stylistic differences. These results highlight the method's potential to quantify and interpret urban style evolution, offering a scientifically grounded lens for future design. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:35:39 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yin",
"Jun",
""
],
[
"Zhong",
"Jing",
""
],
[
"Li",
"Peilin",
""
],
[
"Zeng",
"Pengyu",
""
],
[
"Zhang",
"Miao",
""
],
[
"Luo",
"Ran",
""
],
[
"Lu",
"Shuai",
""
]
] |
2506.10343 | Dongwon Jung | Dongwon Jung, Wenxuan Zhou, Muhao Chen | Code Execution as Grounded Supervision for LLM Reasoning | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Training large language models (LLMs) with chain-of-thought (CoT) supervision has proven effective for enhancing their reasoning abilities. However, obtaining reliable and accurate reasoning supervision remains a significant challenge. We propose a scalable method for generating a high-quality CoT supervision dataset by leveraging the determinism of program execution. Unlike existing reasoning dataset generation methods that rely on costly human annotations or error-prone LLM-generated CoT, our approach extracts verifiable, step-by-step reasoning traces from code execution and transforms them into a natural language CoT reasoning. Experiments on reasoning benchmarks across various domains show that our method effectively equips LLMs with transferable reasoning abilities across diverse tasks. Furthermore, the ablation studies validate that our method produces highly accurate reasoning data and reduces overall token length during inference by reducing meaningless repetition and overthinking. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:36:57 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Jung",
"Dongwon",
""
],
[
"Zhou",
"Wenxuan",
""
],
[
"Chen",
"Muhao",
""
]
] |
2506.10344 | Mina Chookhachizadeh Moghadam Dr | Mina C. Moghadam, Alan Q. Wang, Omer Taub, Martin R. Prince, Mert R. Sabuncu | RealKeyMorph: Keypoints in Real-world Coordinates for Resolution-agnostic Image Registration | 23 pages, 8 figures, to be submitted to MELBA | null | null | null | cs.CV | http://creativecommons.org/publicdomain/zero/1.0/ | Many real-world settings require registration of a pair of medical images that differ in spatial resolution, which may arise from differences in image acquisition parameters like pixel spacing, slice thickness, and field-of-view. However, all previous machine learning-based registration techniques resample images onto a fixed resolution. This is suboptimal because resampling can introduce artifacts due to interpolation. To address this, we present RealKeyMorph (RKM), a resolution-agnostic method for image registration. RKM is an extension of KeyMorph, a registration framework which works by training a network to learn corresponding keypoints for a given pair of images, after which a closed-form keypoint matching step is used to derive the transformation that aligns them. To avoid resampling and enable operating on the raw data, RKM outputs keypoints in real-world coordinates of the scanner. To do this, we leverage the affine matrix produced by the scanner (e.g., MRI machine) that encodes the mapping from voxel coordinates to real world coordinates. By transforming keypoints into real-world space and integrating this into the training process, RKM effectively enables the extracted keypoints to be resolution-agnostic. In our experiments, we demonstrate the advantages of RKM on the registration task for orthogonal 2D stacks of abdominal MRIs, as well as 3D volumes with varying resolutions in brain datasets. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:40:32 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Moghadam",
"Mina C.",
""
],
[
"Wang",
"Alan Q.",
""
],
[
"Taub",
"Omer",
""
],
[
"Prince",
"Martin R.",
""
],
[
"Sabuncu",
"Mert R.",
""
]
] |
2506.10345 | Philipp B\"ar | Philipp B\"ar and Moe T. Wynn and Sander J. J. Leemans | Technical Report with Proofs for A Full Picture in Conformance Checking: Efficiently Summarizing All Optimal Alignments | null | null | null | null | cs.IT cs.LG math.IT | http://creativecommons.org/licenses/by/4.0/ | This technical report provides proofs for the claims in the paper "A Full Picture in Conformance Checking: Efficiently Summarizing All Optimal Alignments". | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:47:06 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Bär",
"Philipp",
""
],
[
"Wynn",
"Moe T.",
""
],
[
"Leemans",
"Sander J. J.",
""
]
] |
2506.10346 | Florian Boudin | Florian Boudin, Akiko Aizawa | An Analysis of Datasets, Metrics and Models in Keyphrase Generation | GEM^2 paper @ ACL 2025 | null | null | null | cs.IR cs.CL | http://creativecommons.org/licenses/by/4.0/ | Keyphrase generation refers to the task of producing a set of words or phrases that summarises the content of a document. Continuous efforts have been dedicated to this task over the past few years, spreading across multiple lines of research, such as model architectures, data resources, and use-case scenarios. Yet, the current state of keyphrase generation remains unknown as there has been no attempt to review and analyse previous work. In this paper, we bridge this gap by presenting an analysis of over 50 research papers on keyphrase generation, offering a comprehensive overview of recent progress, limitations, and open challenges. Our findings highlight several critical issues in current evaluation practices, such as the concerning similarity among commonly-used benchmark datasets and inconsistencies in metric calculations leading to overestimated performances. Additionally, we address the limited availability of pre-trained models by releasing a strong PLM-based model for keyphrase generation as an effort to facilitate future research. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:54:44 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Boudin",
"Florian",
""
],
[
"Aizawa",
"Akiko",
""
]
] |
2506.10347 | Yanhui Li | Yanhui Li, Dongxia Wang, Zhu Sun, Haonan Zhang, Huizhong Guo | LightKG: Efficient Knowledge-Aware Recommendations with Simplified GNN Architecture | Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining | null | 10.1145/3711896.3737026 | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, Graph Neural Networks (GNNs) have become the dominant approach for Knowledge Graph-aware Recommender Systems (KGRSs) due to their proven effectiveness. Building upon GNN-based KGRSs, Self-Supervised Learning (SSL) has been incorporated to address the sparity issue, leading to longer training time. However, through extensive experiments, we reveal that: (1)compared to other KGRSs, the existing GNN-based KGRSs fail to keep their superior performance under sparse interactions even with SSL. (2) More complex models tend to perform worse in sparse interaction scenarios and complex mechanisms, like attention mechanism, can be detrimental as they often increase learning difficulty. Inspired by these findings, we propose LightKG, a simple yet powerful GNN-based KGRS to address sparsity issues. LightKG includes a simplified GNN layer that encodes directed relations as scalar pairs rather than dense embeddings and employs a linear aggregation framework, greatly reducing the complexity of GNNs. Additionally, LightKG incorporates an efficient contrastive layer to implement SSL. It directly minimizes the node similarity in original graph, avoiding the time-consuming subgraph generation and comparison required in previous SSL methods. Experiments on four benchmark datasets show that LightKG outperforms 12 competitive KGRSs in both sparse and dense scenarios while significantly reducing training time. Specifically, it surpasses the best baselines by an average of 5.8\% in recommendation accuracy and saves 84.3\% of training time compared to KGRSs with SSL. Our code is available at https://github.com/1371149/LightKG. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:56:10 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Li",
"Yanhui",
""
],
[
"Wang",
"Dongxia",
""
],
[
"Sun",
"Zhu",
""
],
[
"Zhang",
"Haonan",
""
],
[
"Guo",
"Huizhong",
""
]
] |
2506.10348 | Mykhailo Klymenko Dr | Mykhailo Klymenko and Thong Hoang and Samuel A. Wilkinson and Bahar Goldozian and Suyu Ma and Xiwei Xu and Qinghua Lu and Muhammad Usman and Liming Zhu | Context-Aware Unit Testing for Quantum Subroutines | 28 pages, 9 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Software testing is a critical component of the classical software development lifecycle, and this principle is expected to hold true for quantum software as it evolves toward large-scale production and adherence to industry standards. Developing and testing quantum software presents unique challenges due to the non-deterministic nature of quantum information, the high dimensionality of the underlying Hilbert space, complex hardware noise, and the inherent non-local properties of quantum systems. In this work, we model quantum subroutines as parametrized quantum channels and explore the feasibility of creating practical unit tests using probabilistic assertions, combined with either quantum tomography or statistical tests. To address the computational complexity associated with unit testing in quantum systems, we propose incorporating context-awareness into the testing process. The trade-offs between accuracy, state space coverage, and efficiency associated with the proposed theoretical framework for quantum unit testing have been demonstrated through its application to a simple three-qubit quantum subroutine that prepares a Greenberger-Horne-Zeilinger state, as well as to subroutines within a program implementing Shor's algorithm. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 04:58:56 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Klymenko",
"Mykhailo",
""
],
[
"Hoang",
"Thong",
""
],
[
"Wilkinson",
"Samuel A.",
""
],
[
"Goldozian",
"Bahar",
""
],
[
"Ma",
"Suyu",
""
],
[
"Xu",
"Xiwei",
""
],
[
"Lu",
"Qinghua",
""
],
[
"Usman",
"Muhammad",
""
],
[
"Zhu",
"Liming",
""
]
] |
2506.10349 | Anfeng Xu | Anfeng Xu, Tiantian Feng, Shrikanth Narayanan | Joint ASR and Speaker Role Tagging with Serialized Output Training | Under review | null | null | null | eess.AS cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automatic Speech Recognition systems have made significant progress with large-scale pre-trained models. However, most current systems focus solely on transcribing the speech without identifying speaker roles, a function that is critical for conversational AI. In this work, we investigate the use of serialized output training (SOT) for joint ASR and speaker role tagging. By augmenting Whisper with role-specific tokens and fine-tuning it with SOT, we enable the model to generate role-aware transcriptions in a single decoding pass. We compare the SOT approach against a self-supervised previous baseline method on two real-world conversational datasets. Our findings show that this approach achieves more than 10% reduction in multi-talker WER, demonstrating its feasibility as a unified model for speaker-role aware speech transcription. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:04:53 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Xu",
"Anfeng",
""
],
[
"Feng",
"Tiantian",
""
],
[
"Narayanan",
"Shrikanth",
""
]
] |
2506.10350 | Weibiao Zhao | Weibiao Zhao, Qiucen Wu, Yuanqi Tang, Yu Zhu | Heterogeneous-IRS-Assisted MIMO Systems: Channel Estimation and Beamforming | 30 pages, 8 figures | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Intelligent reflecting surface (IRS) has gained great attention for its ability to create favorable propagation environments. However, the power consumption of conventional IRSs cannot be ignored due to the large number of reflecting elements and control circuits. To balance performance and power consumption, we previously proposed a heterogeneous-IRS (HE-IRS), a green IRS structure integrating dynamically tunable elements (DTEs) and statically tunable elements (STEs). Compared to conventional IRSs with only DTEs, the unique DTE-STE integrated structure introduces new challenges in both channel estimation and beamforming. In this paper, we investigate the channel estimation and beamforming problems in HE-IRS-assisted multi-user multiple-input multiple-output systems. Unlike the overall cascaded channel estimated in conventional IRSs, we show that the HE-IRS channel to be estimated is decomposed into a DTE-based cascaded channel and an STE-based equivalent channel. Leveraging it along with the inherent sparsity of DTE- and STE-based channels and manifold optimization, we propose an efficient channel estimation scheme. To address the rank mismatch problem in the imperfect channel sparsity information, a robust rank selection rule is developed. For beamforming, we propose an offline algorithm to optimize the STE phase shifts for wide beam coverage, and an online algorithm to optimize the BS precoder and the DTE phase shifts using the estimated HE-IRS channel. Simulation results show that the HE-IRS requires less pilot overhead than conventional IRSs with the same number of elements. With the proposed channel estimation and beamforming schemes, the green HE-IRS achieves competitive sum rate performance with significantly reduced power consumption. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:07:41 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhao",
"Weibiao",
""
],
[
"Wu",
"Qiucen",
""
],
[
"Tang",
"Yuanqi",
""
],
[
"Zhu",
"Yu",
""
]
] |
2506.10351 | Yanlong Chen | Yanlong Chen, Mattia Orlandi, Pierangelo Maria Rapa, Simone Benatti, Luca Benini, and Yawei Li | PhysioWave: A Multi-Scale Wavelet-Transformer for Physiological Signal Representation | 22 pages, 8 figures, 9 tables. Submitted to NeurIPS 2025 | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Physiological signals are often corrupted by motion artifacts, baseline drift, and other low-SNR disturbances, which pose significant challenges for analysis. Additionally, these signals exhibit strong non-stationarity, with sharp peaks and abrupt changes that evolve continuously, making them difficult to represent using traditional time-domain or filtering methods. To address these issues, a novel wavelet-based approach for physiological signal analysis is presented, aiming to capture multi-scale time-frequency features in various physiological signals. Leveraging this technique, two large-scale pretrained models specific to EMG and ECG are introduced for the first time, achieving superior performance and setting new baselines in downstream tasks. Additionally, a unified multi-modal framework is constructed by integrating pretrained EEG model, where each modality is guided through its dedicated branch and fused via learnable weighted fusion. This design effectively addresses challenges such as low signal-to-noise ratio, high inter-subject variability, and device mismatch, outperforming existing methods on multi-modal tasks. The proposed wavelet-based architecture lays a solid foundation for analysis of diverse physiological signals, while the multi-modal design points to next-generation physiological signal processing with potential impact on wearable health monitoring, clinical diagnostics, and broader biomedical applications. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:11:41 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chen",
"Yanlong",
""
],
[
"Orlandi",
"Mattia",
""
],
[
"Rapa",
"Pierangelo Maria",
""
],
[
"Benatti",
"Simone",
""
],
[
"Benini",
"Luca",
""
],
[
"Li",
"Yawei",
""
]
] |
2506.10352 | Binyao Guo | Binyao Guo, Zihan Lin, QiZhi He | History-Aware Neural Operator: Robust Data-Driven Constitutive Modeling of Path-Dependent Materials | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study presents an end-to-end learning framework for data-driven modeling of path-dependent inelastic materials using neural operators. The framework is built on the premise that irreversible evolution of material responses, governed by hidden dynamics, can be inferred from observable data.
We develop the History-Aware Neural Operator (HANO), an autoregressive model that predicts path-dependent material responses from short segments of recent strain-stress history without relying on hidden state variables, thereby overcoming self-consistency issues commonly encountered in recurrent neural network (RNN)-based models. Built on a Fourier-based neural operator backbone, HANO enables discretization-invariant learning. To enhance its ability to capture both global loading patterns and critical local path dependencies, we embed a hierarchical self-attention mechanism that facilitates multiscale feature extraction.
Beyond ensuring self-consistency, HANO mitigates sensitivity to initial hidden states, a commonly overlooked issue that can lead to instability in recurrent models when applied to generalized loading paths. By modeling stress-strain evolution as a continuous operator rather than relying on fixed input-output mappings, HANO naturally accommodates varying path discretizations and exhibits robust performance under complex conditions, including irregular sampling, multi-cycle loading, noisy data, and pre-stressed states. We evaluate HANO on two benchmark problems: elastoplasticity with hardening and progressive anisotropic damage in brittle solids. Results show that HANO consistently outperforms baseline models in predictive accuracy, generalization, and robustness. With its demonstrated capabilities, HANO provides an effective data-driven surrogate for simulating inelastic materials and is well-suited for integration with classical numerical solvers. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:19:17 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Guo",
"Binyao",
""
],
[
"Lin",
"Zihan",
""
],
[
"He",
"QiZhi",
""
]
] |
2506.10353 | Runqi Ouyang | Runqi Ouyang and Haoyun Li and Zhenyuan Zhang and Xiaofeng Wang and Zheng Zhu and Guan Huang and Xingang Wang | Motion-R1: Chain-of-Thought Reasoning and Reinforcement Learning for Human Motion Generation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in large language models, especially in natural language understanding and reasoning, have opened new possibilities for text-to-motion generation. Although existing approaches have made notable progress in semantic alignment and motion synthesis, they often rely on end-to-end mapping strategies that fail to capture deep linguistic structures and logical reasoning. Consequently, generated motions tend to lack controllability, consistency, and diversity. To address these limitations, we propose Motion-R1, a unified motion-language modeling framework that integrates a Chain-of-Thought mechanism. By explicitly decomposing complex textual instructions into logically structured action paths, Motion-R1 provides high-level semantic guidance for motion generation, significantly enhancing the model's ability to interpret and execute multi-step, long-horizon, and compositionally rich commands. To train our model, we adopt Group Relative Policy Optimization, a reinforcement learning algorithm designed for large models, which leverages motion quality feedback to optimize reasoning chains and motion synthesis jointly. Extensive experiments across multiple benchmark datasets demonstrate that Motion-R1 achieves competitive or superior performance compared to state-of-the-art methods, particularly in scenarios requiring nuanced semantic understanding and long-term temporal coherence. The code, model and data will be publicly available. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:21:43 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ouyang",
"Runqi",
""
],
[
"Li",
"Haoyun",
""
],
[
"Zhang",
"Zhenyuan",
""
],
[
"Wang",
"Xiaofeng",
""
],
[
"Zhu",
"Zheng",
""
],
[
"Huang",
"Guan",
""
],
[
"Wang",
"Xingang",
""
]
] |
2506.10354 | Reese Pathak | Liviu Aolaritei, Michael I. Jordan, Reese Pathak, Annie Ulichney | Revisiting mean estimation over $\ell_p$ balls: Is the MLE optimal? | 37 pages, 3 figures | null | null | null | math.ST cs.IT math.IT stat.TH | http://creativecommons.org/publicdomain/zero/1.0/ | We revisit the problem of mean estimation on $\ell_p$ balls under additive Gaussian noise. When $p$ is strictly less than $2$, it is well understood that rate-optimal estimators must be nonlinear in the observations. In this work, we study the maximum likelihood estimator (MLE), which may be viewed as a nonlinear shrinkage procedure for mean estimation over $\ell_p$ balls. We demonstrate two phenomena for the behavior of the MLE, which depend on the noise level, the radius of the norm constraint, the dimension, and the norm index $p$. First, as a function of the dimension, for $p$ near $1$ or at least $2$, the MLE is minimax rate-optimal for all noise levels and all constraint radii. On the other hand, for $p$ between $1$ and $2$, there is a more striking behavior: for essentially all noise levels and radii for which nonlinear estimates are required, the MLE is minimax rate-suboptimal, despite being nonlinear in the observations. Our results also imply similar conclusions when given $n$ independent and identically distributed Gaussian samples, where we demonstrate that the MLE can be suboptimal by a polynomial factor in the sample size. Our lower bounds are constructive: whenever the MLE is rate-suboptimal, we provide explicit instances on which the MLE provably incurs suboptimal risk. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:23:01 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Aolaritei",
"Liviu",
""
],
[
"Jordan",
"Michael I.",
""
],
[
"Pathak",
"Reese",
""
],
[
"Ulichney",
"Annie",
""
]
] |
2506.10355 | Yu-Yang Qian | Yu-Yang Qian, Yuan-Ze Xu, Zhen-Yu Zhang, Peng Zhao, and Zhi-Hua Zhou | TreeLoRA: Efficient Continual Learning via Layer-Wise LoRAs Guided by a Hierarchical Gradient-Similarity Tree | ICML 2025 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many real-world applications collect data in a streaming environment, where learning tasks are encountered sequentially. This necessitates continual learning (CL) to update models online, enabling adaptation to new tasks while preserving past knowledge to prevent catastrophic forgetting. Nowadays, with the flourish of large pre-trained models (LPMs), efficiency has become increasingly critical for CL, due to their substantial computational demands and growing parameter sizes. In this paper, we introduce TreeLoRA (K-D Tree of Low-Rank Adapters), a novel approach that constructs layer-wise adapters by leveraging hierarchical gradient similarity to enable efficient CL, particularly for LPMs. To reduce the computational burden of task similarity estimation, we employ bandit techniques to develop an algorithm based on lower confidence bounds to efficiently explore the task structure. Furthermore, we use sparse gradient updates to facilitate parameter optimization, making the approach better suited for LPMs. Theoretical analysis is provided to justify the rationale behind our approach, and experiments on both vision transformers (ViTs) and large language models (LLMs) demonstrate the effectiveness and efficiency of our approach across various domains, including vision and natural language processing tasks. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:25:35 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Qian",
"Yu-Yang",
""
],
[
"Xu",
"Yuan-Ze",
""
],
[
"Zhang",
"Zhen-Yu",
""
],
[
"Zhao",
"Peng",
""
],
[
"Zhou",
"Zhi-Hua",
""
]
] |
2506.10356 | Omid Asudeh | Omid Asudeh, Sina Mahdipour Saravani, Gerald Sabin, Fabrice Rastello, P Sadayappan | Is Sparse Matrix Reordering Effective for Sparse Matrix-Vector Multiplication? | null | null | null | null | cs.DC cs.PF | http://creativecommons.org/licenses/by/4.0/ | This work evaluates the impact of sparse matrix reordering on the performance of sparse matrix-vector multiplication across different multicore CPU platforms. Reordering can significantly enhance performance by optimizing the non-zero element patterns to reduce total data movement and improve the load-balancing. We examine how these gains vary over different CPUs for different reordering strategies, focusing on both sequential and parallel execution. We address multiple aspects, including appropriate measurement methodology, comparison across different kinds of reordering strategies, consistency across machines, and impact of load imbalance. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:29:39 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Asudeh",
"Omid",
""
],
[
"Saravani",
"Sina Mahdipour",
""
],
[
"Sabin",
"Gerald",
""
],
[
"Rastello",
"Fabrice",
""
],
[
"Sadayappan",
"P",
""
]
] |
2506.10357 | Zaijing Li | Zaijing Li, Yuquan Xie, Rui Shao, Gongwei Chen, Weili Guan, Dongmei Jiang, Liqiang Nie | Optimus-3: Towards Generalist Multimodal Minecraft Agents with Scalable Task Experts | 24 pages, 10 figures | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, agents based on multimodal large language models (MLLMs) have achieved remarkable progress across various domains. However, building a generalist agent with capabilities such as perception, planning, action, grounding, and reflection in open-world environments like Minecraft remains challenges: insufficient domain-specific data, interference among heterogeneous tasks, and visual diversity in open-world settings. In this paper, we address these challenges through three key contributions. 1) We propose a knowledge-enhanced data generation pipeline to provide scalable and high-quality training data for agent development. 2) To mitigate interference among heterogeneous tasks, we introduce a Mixture-of-Experts (MoE) architecture with task-level routing. 3) We develop a Multimodal Reasoning-Augmented Reinforcement Learning approach to enhance the agent's reasoning ability for visual diversity in Minecraft. Built upon these innovations, we present Optimus-3, a general-purpose agent for Minecraft. Extensive experimental results demonstrate that Optimus-3 surpasses both generalist multimodal large language models and existing state-of-the-art agents across a wide range of tasks in the Minecraft environment. Project page: https://cybertronagent.github.io/Optimus-3.github.io/ | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:29:40 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Li",
"Zaijing",
""
],
[
"Xie",
"Yuquan",
""
],
[
"Shao",
"Rui",
""
],
[
"Chen",
"Gongwei",
""
],
[
"Guan",
"Weili",
""
],
[
"Jiang",
"Dongmei",
""
],
[
"Nie",
"Liqiang",
""
]
] |
2506.10358 | Satoru Hayami | Ryota Yambe, Satoru Hayami | Symmetry Rules on Multipole Interactions under Crystallographic Point Groups and Application to Multiple-$Q$ Multipole States | 23 pages, 2 figures, 27 tables | J. Phys. Soc. Jpn. 94, 074704 (2025) | 10.7566/JPSJ.94.074704 | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Multipole degrees of freedom describe the mutual interplay among the charge, spin, and orbital degrees of freedom in electrons, which provides a microscopic understanding of unconventional electronic orderings and their associated physical phenomena. We here show the symmetry rules on multipole interactions under crystallographic point groups in a systematic manner. Depending on the bond symmetries, we show the necessary symmetry conditions of the antisymmetric multipole interactions, which correspond to the extension of the Dzyaloshinskii-Moriya interaction, as well as the symmetric ones, which correspond to the extension of the compasslike interaction. Furthermore, we demonstrate that the symmetry-allowed multipole interactions can become a source of exotic multiple-$Q$ multipole orderings. As a specific example, we analyze the effective model with the antisymmetric quadrupole interaction on a triangular lattice and show the emergence of the triple-$Q$ quadrupole state. Our results indicate that multipole interactions that often arise from the heavy-fermion, frustrated, and nematic systems can potentially induce further unconventional quantum states of matter. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:32:45 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yambe",
"Ryota",
""
],
[
"Hayami",
"Satoru",
""
]
] |
2506.10359 | Che Wang | Che Wang, Jeroen van Baar, Chaitanya Mitash, Shuai Li, Dylan Randle, Weiyao Wang, Sumedh Sontakke, Kostas E. Bekris, Kapil Katyal | Demonstrating Multi-Suction Item Picking at Scale via Multi-Modal Learning of Pick Success | Accepted to Robotics: Science and Systems (RSS 2025), 15 pages | null | null | null | cs.RO cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work demonstrates how autonomously learning aspects of robotic operation from sparsely-labeled, real-world data of deployed, engineered solutions at industrial scale can provide with solutions that achieve improved performance. Specifically, it focuses on multi-suction robot picking and performs a comprehensive study on the application of multi-modal visual encoders for predicting the success of candidate robotic picks. Picking diverse items from unstructured piles is an important and challenging task for robot manipulation in real-world settings, such as warehouses. Methods for picking from clutter must work for an open set of items while simultaneously meeting latency constraints to achieve high throughput. The demonstrated approach utilizes multiple input modalities, such as RGB, depth and semantic segmentation, to estimate the quality of candidate multi-suction picks. The strategy is trained from real-world item picking data, with a combination of multimodal pretrain and finetune. The manuscript provides comprehensive experimental evaluation performed over a large item-picking dataset, an item-picking dataset targeted to include partial occlusions, and a package-picking dataset, which focuses on containers, such as boxes and envelopes, instead of unpackaged items. The evaluation measures performance for different item configurations, pick scenes, and object types. Ablations help to understand the effects of in-domain pretraining, the impact of different modalities and the importance of finetuning. These ablations reveal both the importance of training over multiple modalities but also the ability of models to learn during pretraining the relationship between modalities so that during finetuning and inference, only a subset of them can be used as input. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:35:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wang",
"Che",
""
],
[
"van Baar",
"Jeroen",
""
],
[
"Mitash",
"Chaitanya",
""
],
[
"Li",
"Shuai",
""
],
[
"Randle",
"Dylan",
""
],
[
"Wang",
"Weiyao",
""
],
[
"Sontakke",
"Sumedh",
""
],
[
"Bekris",
"Kostas E.",
""
],
[
"Katyal",
"Kapil",
""
]
] |
2506.10360 | Sugilesh H | Ambily A A and Sugilesh H | Horrocks theorem for odd orthogonal groups | null | null | null | null | math.KT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove Horrocks' theorem for the odd elementary orthogonal group, which gives a decomposition of an orthogonal matrix with entries from a polynomial ring $R[X]$, over a commutative ring $R$ in which 2 is invertible, as a product of an orthogonal matrix with entries in $R$ and an elementary orthogonal matrix with entries from $R[X]$. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:35:55 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"A",
"Ambily A",
""
],
[
"H",
"Sugilesh",
""
]
] |
2506.10361 | Novendra Setyawan | Novendra Setyawan, Chi-Chia Sun, Mao-Hsiu Hsu, Wen-Kai Kuo, Jun-Wei Hsieh | FaceLiVT: Face Recognition using Linear Vision Transformer with Structural Reparameterization For Mobile Device | 2025 ICIP | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | This paper introduces FaceLiVT, a lightweight yet powerful face recognition model that integrates a hybrid Convolution Neural Network (CNN)-Transformer architecture with an innovative and lightweight Multi-Head Linear Attention (MHLA) mechanism. By combining MHLA alongside a reparameterized token mixer, FaceLiVT effectively reduces computational complexity while preserving competitive accuracy. Extensive evaluations on challenging benchmarks; including LFW, CFP-FP, AgeDB-30, IJB-B, and IJB-C; highlight its superior performance compared to state-of-the-art lightweight models. MHLA notably improves inference speed, allowing FaceLiVT to deliver high accuracy with lower latency on mobile devices. Specifically, FaceLiVT is 8.6 faster than EdgeFace, a recent hybrid CNN-Transformer model optimized for edge devices, and 21.2 faster than a pure ViT-Based model. With its balanced design, FaceLiVT offers an efficient and practical solution for real-time face recognition on resource-constrained platforms. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:36:40 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Setyawan",
"Novendra",
""
],
[
"Sun",
"Chi-Chia",
""
],
[
"Hsu",
"Mao-Hsiu",
""
],
[
"Kuo",
"Wen-Kai",
""
],
[
"Hsieh",
"Jun-Wei",
""
]
] |
2506.10362 | Yeqing Qiu | Yeqing Qiu, Chengpiao Huang, Ye Xue, Zhipeng Jiang, Qingjiang Shi, Dong Zhang, Zhi-Quan Luo | Relaxation-Free Min-k-Partition for PCI Assignment in 5G Networks | null | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Physical Cell Identity (PCI) is a critical parameter in 5G networks. Efficient and accurate PCI assignment is essential for mitigating mod-3 interference, mod-30 interference, collisions, and confusions among cells, which directly affect network reliability and user experience. In this paper, we propose a novel framework for PCI assignment by decomposing the problem into Min-3-Partition, Min-10-Partition, and a graph coloring problem, leveraging the Chinese Remainder Theorem (CRT). Furthermore, we develop a relaxation-free approach to the general Min-$k$-Partition problem by reformulating it as a quadratic program with a norm-equality constraint and solving it using a penalized mirror descent (PMD) algorithm. The proposed method demonstrates superior computational efficiency and scalability, significantly reducing interference while eliminating collisions and confusions in large-scale 5G networks. Numerical evaluations on real-world datasets show that our approach reduces computational time by up to 20 times compared to state-of-the-art methods, making it highly practical for real-time PCI optimization in large-scale networks. These results highlight the potential of our method to improve network performance and reduce deployment costs in modern 5G systems. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:39:14 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Qiu",
"Yeqing",
""
],
[
"Huang",
"Chengpiao",
""
],
[
"Xue",
"Ye",
""
],
[
"Jiang",
"Zhipeng",
""
],
[
"Shi",
"Qingjiang",
""
],
[
"Zhang",
"Dong",
""
],
[
"Luo",
"Zhi-Quan",
""
]
] |
2506.10363 | Daniel Betschinske | Daniel Betschinske, Malte Schrimpf, Steven Peters, Kamil Klonecki, Jan Peter Karch, Moritz Lippert | Towards more efficient quantitative safety validation of residual risk for assisted and automated driving | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The safety validation of Advanced Driver Assistance Systems (ADAS) and Automated Driving Systems (ADS) increasingly demands efficient and reliable methods to quantify residual risk while adhering to international standards such as ISO 21448. Traditionally, Field Operational Testing (FOT) has been pivotal for macroscopic safety validation of automotive driving functions up to SAE automation level 2. However, state-of-the-art derivations for empirical safety demonstrations using FOT often result in impractical testing efforts, particularly at higher automation levels. Even at lower automation levels, this limitation - coupled with the substantial costs associated with FOT - motivates the exploration of approaches to enhance the efficiency of FOT-based macroscopic safety validation. Therefore, this publication systematically identifies and evaluates state-of-the-art Reduction Approaches (RAs) for FOT, including novel methods reported in the literature. Based on an analysis of ISO 21448, two models are derived: a generic model capturing the argumentation components of the standard, and a base model, exemplarily applied to Automatic Emergency Braking (AEB) systems, establishing a baseline for the real-world driving requirement for a Quantitative Safety Validation of Residual Risk (QSVRR). Subsequently, the RAs are assessed using four criteria: quantifiability, threats to validity, missing links, and black box compatibility, highlighting potential benefits, inherent limitations, and identifying key areas for further research. Our evaluation reveals that, while several approaches offer potential, none are free from missing links or other substantial shortcomings. Moreover, no identified alternative can fully replace FOT, reflecting its crucial role in the safety validation of ADAS and ADS. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:41:53 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Betschinske",
"Daniel",
""
],
[
"Schrimpf",
"Malte",
""
],
[
"Peters",
"Steven",
""
],
[
"Klonecki",
"Kamil",
""
],
[
"Karch",
"Jan Peter",
""
],
[
"Lippert",
"Moritz",
""
]
] |
2506.10364 | Pengrun Huang | Penguin Huang, Chhavi Yadav, Ruihan Wu, Kamalika Chaudhuri | Can We Infer Confidential Properties of Training Data from LLMs? | null | null | null | null | cs.LG cs.CL cs.CR | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) are increasingly fine-tuned on domain-specific datasets to support applications in fields such as healthcare, finance, and law. These fine-tuning datasets often have sensitive and confidential dataset-level properties -- such as patient demographics or disease prevalence -- that are not intended to be revealed. While prior work has studied property inference attacks on discriminative models (e.g., image classification models) and generative models (e.g., GANs for image data), it remains unclear if such attacks transfer to LLMs. In this work, we introduce PropInfer, a benchmark task for evaluating property inference in LLMs under two fine-tuning paradigms: question-answering and chat-completion. Built on the ChatDoctor dataset, our benchmark includes a range of property types and task configurations. We further propose two tailored attacks: a prompt-based generation attack and a shadow-model attack leveraging word frequency signals. Empirical evaluations across multiple pretrained LLMs show the success of our attacks, revealing a previously unrecognized vulnerability in LLMs. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:42:06 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Huang",
"Penguin",
""
],
[
"Yadav",
"Chhavi",
""
],
[
"Wu",
"Ruihan",
""
],
[
"Chaudhuri",
"Kamalika",
""
]
] |
2506.10365 | Shuyang Hou | Shuyang Hou, Zhangxiao Shen, Huayi Wu, Haoyue Jiao, Ziqi Liu, Lutong Xie, Chang Liu, Jianyuan Liang, Yaxian Qing, Xiaopu Zhang, Dehua Peng, Zhipeng Gui, Xuefeng Guan | AutoGEEval++: A Multi-Level and Multi-Geospatial-Modality Automated Evaluation Framework for Large Language Models in Geospatial Code Generation on Google Earth Engine | null | null | null | null | cs.SE | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Geospatial code generation is becoming a key frontier in integrating artificial intelligence with geo-scientific analysis, yet standardised automated evaluation tools for this task remain absent. This study presents AutoGEEval++, an enhanced framework building on AutoGEEval, and the first automated assessment system for large language models (LLMs) generating geospatial code on Google Earth Engine (GEE). It supports diverse data modalities and varying task complexities. Built on the GEE Python API, AutoGEEval++ features a benchmark dataset-AutoGEEval++-Bench-with 6,365 test cases across 26 data types and three task categories: unit, combo, and theme tests. It includes a submission programme and a judge module to realise an end-to-end automated evaluation pipeline from code generation to execution-based validation. The framework adopts multi-dimensional metrics-accuracy, resource usage, run-time efficiency, and error types-balancing hallucination control and efficiency, and enabling boundary testing and error pattern analysis. Using AutoGEEval++, we evaluate 24 state-of-the-art LLMs (as of June 2025), including general-purpose, reasoning-enhanced, code-centric, and geoscience-specific models. Results reveal clear performance, stability, and error differences across task types, model designs, and deployment settings, confirming AutoGEEval++'s practical value and scalability in vertical-domain code generation. This work establishes the first standardised evaluation protocol and foundational benchmark for GEE-based LLM code generation, providing a unified basis for performance comparison and a methodological framework for systematic, domain-specific code evaluation. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:42:37 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Hou",
"Shuyang",
""
],
[
"Shen",
"Zhangxiao",
""
],
[
"Wu",
"Huayi",
""
],
[
"Jiao",
"Haoyue",
""
],
[
"Liu",
"Ziqi",
""
],
[
"Xie",
"Lutong",
""
],
[
"Liu",
"Chang",
""
],
[
"Liang",
"Jianyuan",
""
],
[
"Qing",
"Yaxian",
""
],
[
"Zhang",
"Xiaopu",
""
],
[
"Peng",
"Dehua",
""
],
[
"Gui",
"Zhipeng",
""
],
[
"Guan",
"Xuefeng",
""
]
] |
2506.10366 | Tianpei Zhang | Tianpei Zhang and Jufeng Zhao and Yiming Zhu and Guangmang Cui and Yuhan Lyu | FSATFusion: Frequency-Spatial Attention Transformer for Infrared and Visible Image Fusion | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The infrared and visible images fusion (IVIF) is receiving increasing attention from both the research community and industry due to its excellent results in downstream applications. Existing deep learning approaches often utilize convolutional neural networks to extract image features. However, the inherently capacity of convolution operations to capture global context can lead to information loss, thereby restricting fusion performance. To address this limitation, we propose an end-to-end fusion network named the Frequency-Spatial Attention Transformer Fusion Network (FSATFusion). The FSATFusion contains a frequency-spatial attention Transformer (FSAT) module designed to effectively capture discriminate features from source images. This FSAT module includes a frequency-spatial attention mechanism (FSAM) capable of extracting significant features from feature maps. Additionally, we propose an improved Transformer module (ITM) to enhance the ability to extract global context information of vanilla Transformer. We conducted both qualitative and quantitative comparative experiments, demonstrating the superior fusion quality and efficiency of FSATFusion compared to other state-of-the-art methods. Furthermore, our network was tested on two additional tasks without any modifications, to verify the excellent generalization capability of FSATFusion. Finally, the object detection experiment demonstrated the superiority of FSATFusion in downstream visual tasks. Our code is available at https://github.com/Lmmh058/FSATFusion. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:42:38 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Tianpei",
""
],
[
"Zhao",
"Jufeng",
""
],
[
"Zhu",
"Yiming",
""
],
[
"Cui",
"Guangmang",
""
],
[
"Lyu",
"Yuhan",
""
]
] |
2506.10367 | Zhi Lei Zhang | Zhi-Lei Zhang, Guo-Jian Qiao and C. P. Sun | Poor Man's Majoranon in Two Quantum Dots Dressed by Superconducting Quasi-Excitations | null | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | In a hybrid system consisting of two quantum dots (QDs) coupled to a superconductor (SC), zero-bias peaks in the differential conductance spectrum have been reported as potential signatures of Majorana fermions (MFs). However, such signatures typically appear only at specific parameter values of the QDs--so-called `sweet spots'--and are referred to as the Poor Man's Majorana (PMM). To investigate whether these signatures can be conclusively attributed to genuine MFs emerging over a continuous parameter range, we present an alternative approach that microscopically incorporates the superconducting effects into the QDs, rather than simply attribute them into two phenomenological parameters of QDs. This forms the dressed Majorana fermions (DMFs), which can be viewed as superpositions of quasi-excitations from both the QDs and the SC. We show that DMFs can localize at one end of a one-dimensional SC and persist across a continuous parameter range, thereby enhancing the feasibility of experimental detection. Our results provide a more accurate description of the PMM in such hybrid systems and offer practical guidance for observing end-localized PMM modes in continuous one-dimensional SC. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:45:33 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Zhi-Lei",
""
],
[
"Qiao",
"Guo-Jian",
""
],
[
"Sun",
"C. P.",
""
]
] |
2506.10368 | Tony Puthenpurakal | Tony J. Puthenpurakal | On lengths of modules over certain Artinian complete intersections | arXiv admin note: substantial text overlap with arXiv:2307.16132 | null | null | null | math.AC | http://creativecommons.org/licenses/by/4.0/ | Let $(Q,\mathfrak{n})$ be a regular local ring of dimension $c \geq 2$ with algebraically closed residue field $k = Q/\mathfrak{n}$.
Let $f_1, f_2, \ldots f_{c-1}, g$ be a regular sequence in $Q$ such that $ f_i \in \mathfrak{n}^2$ for all $i$ and $g \in \mathfrak{n}$.
Set $A = Q/(f_1,\ldots, f_{c-1}, g^r)$ with $r \geq 2$. Notice $A$ is an Artinian complete intersection of codimension $c$.
We show that there exists $\alpha_A \in \mathbb{P}^{c-1}(k)$ such that there exists integer $m_A \geq 2$ (depending only on $A$) with $m_A$ dividing $\ell(M)$ for every finitely generated $A$-module $M$ with $\alpha_A \notin \mathcal{V}(M)$ (here $\ell(M)$ denotes the length of $M$ and $\mathcal{V}(M)$ denotes the support variety of $M$). As an application we prove that if $k$ be a field and $R = k[X_1, \ldots, X_c]/(X_1^{a_1}, \ldots, X_c^{a_c})$ with $a_i \geq 2$ and $c \geq 2$. Let $p$ be a prime number and assume $p$ divides two of the $a_i$. Then $p$ divides $\ell(E)$ for any $A$-module with bounded betti numbers. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:45:57 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Puthenpurakal",
"Tony J.",
""
]
] |
2506.10369 | Wishnu Badrawani | Wishnu Badrawani | An Interpretable Machine Learning Approach in Predicting Inflation Using Payments System Data: A Case Study of Indonesia | 26 pages, 2 tables, 13 figures | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | This paper evaluates the performance of prominent machine learning (ML) algorithms in predicting Indonesia's inflation using the payment system, capital market, and macroeconomic data. We compare the forecasting performance of each ML model, namely shrinkage regression, ensemble learning, and super vector regression, to that of the univariate time series ARIMA and SARIMA models. We examine various out-of-bag sample periods in each ML model to determine the appropriate data-splitting ratios for the regression case study. This study indicates that all ML models produced lower RMSEs and reduced average forecast errors by 45.16 percent relative to the ARIMA benchmark, with the Extreme Gradient Boosting model outperforming other ML models and the benchmark. Using the Shapley value, we discovered that numerous payment system variables significantly predict inflation. We explore the ML forecast using local Shapley decomposition and show the relationship between the explanatory variables and inflation for interpretation. The interpretation of the ML forecast highlights some significant findings and offers insightful recommendations, enhancing previous economic research that uses a more established econometric method. Our findings advocate ML models as supplementary tools for the central bank to predict inflation and support monetary policy. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:46:15 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Badrawani",
"Wishnu",
""
]
] |
2506.10370 | Xiaodong Li | Xiaohan Hu, Zhentao Li, Xiaodong Li | Estimating Signal-to-Noise Ratios for Multivariate High-dimensional Linear Models | 79 pages, 5 tables | null | null | null | math.ST stat.ME stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Signal-to-noise ratios (SNR) play a crucial role in various statistical models, with important applications in tasks such as estimating heritability in genomics. The method-of-moments estimator is a widely used approach for estimating SNR, primarily explored in single-response settings. In this study, we extend the method-of-moments SNR estimation framework to encompass both fixed effects and random effects linear models with multivariate responses. In particular, we establish and compare the asymptotic distributions of the proposed estimators. Furthermore, we extend our approach to accommodate cases with residual heteroskedasticity and derive asymptotic inference procedures based on standard error estimation. The effectiveness of our methods is demonstrated through extensive numerical experiments. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:46:40 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Hu",
"Xiaohan",
""
],
[
"Li",
"Zhentao",
""
],
[
"Li",
"Xiaodong",
""
]
] |
2506.10371 | Laziz Abdullaev | Laziz U. Abdullaev, Maksim Tkachenko, Tan M. Nguyen | Revisiting Transformers with Insights from Image Filtering | 12 pages, 6 figures | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | The self-attention mechanism, a cornerstone of Transformer-based state-of-the-art deep learning architectures, is largely heuristic-driven and fundamentally challenging to interpret. Establishing a robust theoretical foundation to explain its remarkable success and limitations has therefore become an increasingly prominent focus in recent research. Some notable directions have explored understanding self-attention through the lens of image denoising and nonparametric regression. While promising, existing frameworks still lack a deeper mechanistic interpretation of various architectural components that enhance self-attention, both in its original formulation and subsequent variants. In this work, we aim to advance this understanding by developing a unifying image processing framework, capable of explaining not only the self-attention computation itself but also the role of components such as positional encoding and residual connections, including numerous later variants. We also pinpoint potential distinctions between the two concepts building upon our framework, and make effort to close this gap. We introduce two independent architectural modifications within transformers. While our primary objective is interpretability, we empirically observe that image processing-inspired modifications can also lead to notably improved accuracy and robustness against data contamination and adversaries across language and vision tasks as well as better long sequence understanding. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:46:57 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Abdullaev",
"Laziz U.",
""
],
[
"Tkachenko",
"Maksim",
""
],
[
"Nguyen",
"Tan M.",
""
]
] |
2506.10372 | Qianyun Wang | Qianyun Wang, Yuan Yuan, Xu Zhang | Sobolev regularity for the $\bar{\partial}$--Neumann operator and transverse vector fields | null | null | null | null | math.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | On a smooth, bounded, pseudoconvex domain in $\mathbb{C}^n$ with $n >2$, inspired by the compactness conditions introduced by Yue Zhang, we present new sufficient conditions for the exact regularity of the $\overline{\partial}$--Neumann operator. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:50:47 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wang",
"Qianyun",
""
],
[
"Yuan",
"Yuan",
""
],
[
"Zhang",
"Xu",
""
]
] |
2506.10373 | Jiajun Hu | Jiajun Hu, Chetan Choppali Sudarshan, Vidya A. Chhabria, Aman Arora | CarbonSet: A Dataset to Analyze Trends and Benchmark the Sustainability of CPUs and GPUs | null | null | null | null | cs.AR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Over the years, the chip industry has consistently developed high-performance processors to address the increasing demands across diverse applications. However, the rapid expansion of chip production has significantly increased carbon emissions, raising critical concerns about environmental sustainability. While researchers have previously modeled the carbon footprint (CFP) at both system and processor levels, a holistic analysis of sustainability trends encompassing the entire chip lifecycle remains lacking. This paper presents CarbonSet, a comprehensive dataset integrating sustainability and performance metrics for CPUs and GPUs over the past decade. CarbonSet aims to benchmark and assess the design of next-generation processors. Leveraging this dataset, we conducted detailed analysis of flagship processors' sustainability trends over the last decade. This paper further highlights that modern processors are not yet sustainably designed, with total carbon emissions increasing more than 50$\times$ in the past three years due to the surging demand driven by the AI boom. Power efficiency remains a significant concern, while advanced process nodes pose new challenges requiring to effectively amortize the dramatically increased manufacturing carbon emissions. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:52:16 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Hu",
"Jiajun",
""
],
[
"Sudarshan",
"Chetan Choppali",
""
],
[
"Chhabria",
"Vidya A.",
""
],
[
"Arora",
"Aman",
""
]
] |
2506.10374 | Daniel McMorrow | Daniel McMorrow and Jonathan Scarlett | Optimal Non-Adaptive Group Testing with One-Sided Error Guarantees | null | null | null | null | cs.IT math.IT math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The group testing problem consists of determining a sparse subset of defective items from within a larger set of items via a series of tests, where each test outcome indicates whether at least one defective item is included in the test. We study the approximate recovery setting, where the recovery criterion of the defective set is relaxed to allow a small number of items to be misclassified. In particular, we consider one-sided approximate recovery criteria, where we allow either only false negative or only false positive misclassifications. Under false negatives only (i.e., finding a subset of defectives), we show that there exists an algorithm matching the optimal threshold of two-sided approximate recovery. Under false positives only (i.e., finding a superset of the defectives), we provide a converse bound showing that the better of two existing algorithms is optimal. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:53:53 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"McMorrow",
"Daniel",
""
],
[
"Scarlett",
"Jonathan",
""
]
] |
2506.10375 | Zurab Silagadze | Alina E. Sagaydak, Zurab K. Silagadze | Asymptotic potential of a rose-shaped disk | 10 pages, 1 figure | Journal of Electrostatics 136 (2025), 104099 | 10.1016/j.elstat.2025.104099 | null | physics.class-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Based on extensive numerical evidence, a recent paper (Sheng Chen et al., Eur. J. Phys. {\bf 45} (2024), 045703) suggested that the potential of a uniformly charged rose-shaped disk in the plane of the disk has a simple asymptotic form. We present an analytical proof of this interesting conjecture. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 05:57:08 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Sagaydak",
"Alina E.",
""
],
[
"Silagadze",
"Zurab K.",
""
]
] |
2506.10376 | Fan Wu | Fan Wu, Cuiyun Gao, Shuqing Li, Xin-Cheng Wen, Qing Liao | MLLM-Based UI2Code Automation Guided by UI Layout Information | Accepted by the 34th International Symposium on Software Testing and Analysis (ISSTA 2025) | null | 10.1145/3728925 | null | cs.SE cs.HC | http://creativecommons.org/licenses/by/4.0/ | Converting user interfaces into code (UI2Code) is a crucial step in website development, which is time-consuming and labor-intensive. The automation of UI2Code is essential to streamline this task, beneficial for improving the development efficiency. There exist deep learning-based methods for the task; however, they heavily rely on a large amount of labeled training data and struggle with generalizing to real-world, unseen web page designs. The advent of Multimodal Large Language Models (MLLMs) presents potential for alleviating the issue, but they are difficult to comprehend the complex layouts in UIs and generate the accurate code with layout preserved. To address these issues, we propose LayoutCoder, a novel MLLM-based framework generating UI code from real-world webpage images, which includes three key modules: (1) Element Relation Construction, which aims at capturing UI layout by identifying and grouping components with similar structures; (2) UI Layout Parsing, which aims at generating UI layout trees for guiding the subsequent code generation process; and (3) Layout-Guided Code Fusion, which aims at producing the accurate code with layout preserved. For evaluation, we build a new benchmark dataset which involves 350 real-world websites named Snap2Code, divided into seen and unseen parts for mitigating the data leakage issue, besides the popular dataset Design2Code. Extensive evaluation shows the superior performance of LayoutCoder over the state-of-the-art approaches. Compared with the best-performing baseline, LayoutCoder improves 10.14% in the BLEU score and 3.95% in the CLIP score on average across all datasets. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:04:16 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wu",
"Fan",
""
],
[
"Gao",
"Cuiyun",
""
],
[
"Li",
"Shuqing",
""
],
[
"Wen",
"Xin-Cheng",
""
],
[
"Liao",
"Qing",
""
]
] |
2506.10377 | Yun Chen Tsai | Yun Chen Tsai and Kittiphon Phalakarn and S. Akshay and Ichiro Hasuo | Chance and Mass Interpretations of Probabilities in Markov Decision Processes (Extended Version) | null | null | null | null | cs.FL | http://creativecommons.org/licenses/by/4.0/ | Markov decision processes (MDPs) are a popular model for decision-making in the presence of uncertainty. The conventional view of MDPs in verification treats them as state transformers with probabilities defined over sequences of states and with schedulers making random choices. An alternative view, especially well-suited for modeling dynamical systems, defines MDPs as distribution transformers with schedulers distributing probability masses. Our main contribution is a unified semantical framework that accommodates these two views and two new ones. These four semantics of MDPs arise naturally through identifying different sources of randomness in an MDP (namely schedulers, configurations, and transitions) and providing different ways of interpreting these probabilities (called the chance and mass interpretations). These semantics are systematically unified through a mathematical construct called chance-mass (CM) classifier. As another main contribution, we study a reachability problem in each of the two new semantics, demonstrating their hardness and providing two algorithms for solving them. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:05:38 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Tsai",
"Yun Chen",
""
],
[
"Phalakarn",
"Kittiphon",
""
],
[
"Akshay",
"S.",
""
],
[
"Hasuo",
"Ichiro",
""
]
] |
2506.10378 | Jikai Jin | Jikai Jin and Vasilis Syrgkanis and Sham Kakade and Hanlin Zhang | Discovering Hierarchical Latent Capabilities of Language Models via Causal Representation Learning | null | null | null | null | cs.LG cs.AI cs.CL stat.ML | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Faithful evaluation of language model capabilities is crucial for deriving actionable insights that can inform model development. However, rigorous causal evaluations in this domain face significant methodological challenges, including complex confounding effects and prohibitive computational costs associated with extensive retraining. To tackle these challenges, we propose a causal representation learning framework wherein observed benchmark performance is modeled as a linear transformation of a few latent capability factors. Crucially, these latent factors are identified as causally interrelated after appropriately controlling for the base model as a common confounder. Applying this approach to a comprehensive dataset encompassing over 1500 models evaluated across six benchmarks from the Open LLM Leaderboard, we identify a concise three-node linear causal structure that reliably explains the observed performance variations. Further interpretation of this causal structure provides substantial scientific insights beyond simple numerical rankings: specifically, we reveal a clear causal direction starting from general problem-solving capabilities, advancing through instruction-following proficiency, and culminating in mathematical reasoning ability. Our results underscore the essential role of carefully controlling base model variations during evaluation, a step critical to accurately uncovering the underlying causal relationships among latent model capabilities. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:07:42 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Jin",
"Jikai",
""
],
[
"Syrgkanis",
"Vasilis",
""
],
[
"Kakade",
"Sham",
""
],
[
"Zhang",
"Hanlin",
""
]
] |
2506.10379 | Xin Wang | Jie Liu and Xin Wang | Hamiltonian Learning via Inverse Physics-Informed Neural Networks | 13 pages, 10 figures | null | null | null | quant-ph cond-mat.dis-nn physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hamiltonian learning (HL), enabling precise estimation of system parameters and underlying dynamics, plays a critical role in characterizing quantum systems. However, conventional HL methods face challenges in noise robustness and resource efficiency, especially under limited measurements. In this work, we present \textit{Inverse Physics-Informed Neural Networks for Hamiltonian Learning (iPINN-HL)}, an approach that embeds the Schr\"{o}dinger equation directly into the machine learning procedure. This formulation allows the model to integrate both observational data and known physical laws to infer Hamiltonian parameters with greater accuracy and resource efficiency. We benchmark iPINN-HL against a deep-neural-network-based quantum state tomography method (denoted as DNN-HL) and demonstrate its effectiveness across several different scenarios, including one-dimensional spin chains, cross-resonance gate calibration, crosstalk identification, and real-time compensation to parameter drift. Our results show that iPINN-HL can approach the Heisenberg limit in certain settings and exhibits robustness to noises, while outperforming DNN-HL in accuracy and resource efficiency. Therefore, iPINN-HL is a powerful and flexible framework for quantum system characterization for practical tasks. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:09:41 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Liu",
"Jie",
""
],
[
"Wang",
"Xin",
""
]
] |
2506.10380 | Xiaohan Yu | Xiaohan Yu, Pu Jian, Chong Chen | TableRAG: A Retrieval Augmented Generation Framework for Heterogeneous Document Reasoning | Under review. Codes are available at https://github.com/yxh-y/TableRAG/tree/main | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | Retrieval-Augmented Generation (RAG) has demonstrated considerable effectiveness in open-domain question answering. However, when applied to heterogeneous documents, comprising both textual and tabular components, existing RAG approaches exhibit critical limitations. The prevailing practice of flattening tables and chunking strategies disrupts the intrinsic tabular structure, leads to information loss, and undermines the reasoning capabilities of LLMs in multi-hop, global queries. To address these challenges, we propose TableRAG, an hybrid framework that unifies textual understanding and complex manipulations over tabular data. TableRAG iteratively operates in four steps: context-sensitive query decomposition, text retrieval, SQL programming and execution, and compositional intermediate answer generation. We also develop HeteQA, a novel benchmark designed to evaluate the multi-hop heterogeneous reasoning capabilities. Experimental results demonstrate that TableRAG consistently outperforms existing baselines on both public datasets and our HeteQA, establishing a new state-of-the-art for heterogeneous document question answering. We release TableRAG at https://github.com/yxh-y/TableRAG/tree/main. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:16:49 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yu",
"Xiaohan",
""
],
[
"Jian",
"Pu",
""
],
[
"Chen",
"Chong",
""
]
] |
2506.10381 | Edgar Martinez-Moro | Sanjit Bhowmick and Kuntal Deka and Alexandre Fotue Tabue and Edgar Mart\'inez-Moro | Trace duality and additive complementary pairs of additive cyclic codes over finite chain rings | null | null | null | null | cs.IT math.IT | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper investigates the algebraic structure of additive complementary pairs of cyclic codes over a finite commutative ring. We demonstrate that for every additive complementary pair of additive cyclic codes, both constituent codes are free modules. Moreover, we present a necessary and sufficient condition for a pair of additive cyclic codes over a finite commutative ring to form an additive complementary pair. Finally, we construct a complementary pair of additive cyclic codes over a finite chain ring and show that one of the codes is permutation equivalent to the trace dual of the other. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:18:16 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Bhowmick",
"Sanjit",
""
],
[
"Deka",
"Kuntal",
""
],
[
"Tabue",
"Alexandre Fotue",
""
],
[
"Martínez-Moro",
"Edgar",
""
]
] |
2506.10382 | Huanchen Bao | Huanchen Bao and Jeff York Ye | Upper cluster structure on Kac--Moody Richardson varieties | 30 pages | null | null | null | math.RT math.AG math.CO | http://creativecommons.org/licenses/by/4.0/ | We show coordinate rings of open Richardson varieties are upper cluster algebras for any symmetrizable Kac--Moody type. We further show the coordinate rings of (generalized) open Richardson varieties on the twisted product of flag varieties are upper cluster algebras for any symmetrizable Kac--Moody type. This includes, as special cases, reduced double Bruhat cells, Bott-Samelson varieties, braid varieties. Our results generalize various results by Casals--Gorsky--Gorsky--Le--Shen--Simental and Galashin--Lam--Sherman-Bennett--Speyer in finite types. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:18:57 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Bao",
"Huanchen",
""
],
[
"Ye",
"Jeff York",
""
]
] |
2506.10383 | Nidhi Parayil | Nidhi Homey Parayil, Thierry Peynot and Chris Lehnert | RICE: Reactive Interaction Controller for Cluttered Canopy Environment | This work has been submitted to the IEEE RAL for possible publication | null | null | null | cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Robotic navigation in dense, cluttered environments such as agricultural canopies presents significant challenges due to physical and visual occlusion caused by leaves and branches. Traditional vision-based or model-dependent approaches often fail in these settings, where physical interaction without damaging foliage and branches is necessary to reach a target. We present a novel reactive controller that enables safe navigation for a robotic arm in a contact-rich, cluttered, deformable environment using end-effector position and real-time tactile feedback. Our proposed framework's interaction strategy is based on a trade-off between minimizing disturbance by maneuvering around obstacles and pushing through them to move towards the target. We show that over 35 trials in 3 experimental plant setups with an occluded target, the proposed controller successfully reached the target in all trials without breaking any branch and outperformed the state-of-the-art model-free controller in robustness and adaptability. This work lays the foundation for safe, adaptive interaction in cluttered, contact-rich deformable environments, enabling future agricultural tasks such as pruning and harvesting in plant canopies. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:19:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Parayil",
"Nidhi Homey",
""
],
[
"Peynot",
"Thierry",
""
],
[
"Lehnert",
"Chris",
""
]
] |
2506.10384 | Yeonghun Lee | Jim O'Connor, Yeonghun Lee, and Gary B Parker | NeuroPAL: Punctuated Anytime Learning with Neuroevolution for Macromanagement in Starcraft: Brood War | IEEE Conference on Games 2025 | null | null | null | cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | StarCraft: Brood War remains a challenging benchmark for artificial intelligence research, particularly in the domain of macromanagement, where long-term strategic planning is required. Traditional approaches to StarCraft AI rely on rule-based systems or supervised deep learning, both of which face limitations in adaptability and computational efficiency. In this work, we introduce NeuroPAL, a neuroevolutionary framework that integrates Neuroevolution of Augmenting Topologies (NEAT) with Punctuated Anytime Learning (PAL) to improve the efficiency of evolutionary training. By alternating between frequent, low-fidelity training and periodic, high-fidelity evaluations, PAL enhances the sample efficiency of NEAT, enabling agents to discover effective strategies in fewer training iterations. We evaluate NeuroPAL in a fixed-map, single-race scenario in StarCraft: Brood War and compare its performance to standard NEAT-based training. Our results show that PAL significantly accelerates the learning process, allowing the agent to reach competitive levels of play in approximately half the training time required by NEAT alone. Additionally, the evolved agents exhibit emergent behaviors such as proxy barracks placement and defensive building optimization, strategies commonly used by expert human players. These findings suggest that structured evaluation mechanisms like PAL can enhance the scalability and effectiveness of neuroevolution in complex real-time strategy environments. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:19:27 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"O'Connor",
"Jim",
""
],
[
"Lee",
"Yeonghun",
""
],
[
"Parker",
"Gary B",
""
]
] |
2506.10385 | Dong-Gil Im | Jungmo Lee, Kyungdeuk Park, Dongkyu Kim, Yonggi Jo, Dong-Gil Im, and Yong Sup Ihn | Optimizing brightness of SPDC source in Laguerre-Gaussian modes using type-0 periodically-poled nonlinear crystal | 14 pages, 7 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Photon pairs generated via spontaneous parametric down-conversion (SPDC) can exhibit entanglement in the Laguerre-Gaussian (LG) mode basis, which enables high-dimensional free-space quantum communication by exploiting the high-dimensional space spanned by the LG modes. For such free-space quantum communication, the brightness of the quantum light source plays an important role due to the atmospheric turbulence and photon loss. A variety of studies have analyzed the SPDC brightness by decomposing biphoton states into LG modes, but they have often relied on a degenerate state, a narrow spectral bandwidth approximation, or a thin crystal approximation. However, these approaches are unsuitable for non-degenerate type-0 SPDC with a periodicallypoled nonlinear crystal, which offers higher brightness due to its superior nonlinear coefficients. In this study, we examine the spectrum of photon pairs in specific LG modes generated by a type-0 ppKTP crystal whileavoiding the constraints imposed by the aforementioned assumptions. In addition, we investigate the optimal focal parameters of the pump, signal, and idler to maximize the brightness for a given LG mode. Our findings show that it is not feasible to simultaneously optimize the brightness for different LG modes with a single pump focal parameter. The results of this study provide a comprehensive framework for developing highbrightness quantum light sources and contribute to the advancement of high-dimensional free-space quantum communication. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:20:00 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lee",
"Jungmo",
""
],
[
"Park",
"Kyungdeuk",
""
],
[
"Kim",
"Dongkyu",
""
],
[
"Jo",
"Yonggi",
""
],
[
"Im",
"Dong-Gil",
""
],
[
"Ihn",
"Yong Sup",
""
]
] |
2506.10386 | Jerry Yan | Jerry Yan, Chinmay Talegaonkar, Nicholas Antipa, Eric Terrill, Sophia Merrifield | Leveraging 6DoF Pose Foundation Models For Mapping Marine Sediment Burial | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The burial state of anthropogenic objects on the seafloor provides insight into localized sedimentation dynamics and is also critical for assessing ecological risks, potential pollutant transport, and the viability of recovery or mitigation strategies for hazardous materials such as munitions. Accurate burial depth estimation from remote imagery remains difficult due to partial occlusion, poor visibility, and object degradation. This work introduces a computer vision pipeline, called PoseIDON, which combines deep foundation model features with multiview photogrammetry to estimate six degrees of freedom object pose and the orientation of the surrounding seafloor from ROV video. Burial depth is inferred by aligning CAD models of the objects with observed imagery and fitting a local planar approximation of the seafloor. The method is validated using footage of 54 objects, including barrels and munitions, recorded at a historic ocean dumpsite in the San Pedro Basin. The model achieves a mean burial depth error of approximately 10 centimeters and resolves spatial burial patterns that reflect underlying sediment transport processes. This approach enables scalable, non-invasive mapping of seafloor burial and supports environmental assessment at contaminated sites. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:21:00 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yan",
"Jerry",
""
],
[
"Talegaonkar",
"Chinmay",
""
],
[
"Antipa",
"Nicholas",
""
],
[
"Terrill",
"Eric",
""
],
[
"Merrifield",
"Sophia",
""
]
] |
2506.10387 | Zaijing Li | Yuquan Xie, Zaijing Li, Rui Shao, Gongwei Chen, Kaiwen Zhou, Yinchuan Li, Dongmei Jiang, Liqiang Nie | Mirage-1: Augmenting and Updating GUI Agent with Hierarchical Multimodal Skills | 20 pages, 5 figures, 5 tables | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent efforts to leverage the Multi-modal Large Language Model (MLLM) as GUI agents have yielded promising outcomes. However, these agents still struggle with long-horizon tasks in online environments, primarily due to insufficient knowledge and the inherent gap between offline and online domains. In this paper, inspired by how humans generalize knowledge in open-ended environments, we propose a Hierarchical Multimodal Skills (HMS) module to tackle the issue of insufficient knowledge. It progressively abstracts trajectories into execution skills, core skills, and ultimately meta-skills, providing a hierarchical knowledge structure for long-horizon task planning. To bridge the domain gap, we propose the Skill-Augmented Monte Carlo Tree Search (SA-MCTS) algorithm, which efficiently leverages skills acquired in offline environments to reduce the action search space during online tree exploration. Building on HMS, we propose Mirage-1, a multimodal, cross-platform, plug-and-play GUI agent. To validate the performance of Mirage-1 in real-world long-horizon scenarios, we constructed a new benchmark, AndroidLH. Experimental results show that Mirage-1 outperforms previous agents by 32\%, 19\%, 15\%, and 79\% on AndroidWorld, MobileMiniWob++, Mind2Web-Live, and AndroidLH, respectively. Project page: https://cybertronagent.github.io/Mirage-1.github.io/ | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:21:19 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Xie",
"Yuquan",
""
],
[
"Li",
"Zaijing",
""
],
[
"Shao",
"Rui",
""
],
[
"Chen",
"Gongwei",
""
],
[
"Zhou",
"Kaiwen",
""
],
[
"Li",
"Yinchuan",
""
],
[
"Jiang",
"Dongmei",
""
],
[
"Nie",
"Liqiang",
""
]
] |
2506.10388 | Stefanie Sonner | Radoslaw Czaja, Stefanie Sonner | A panoramic view of exponential attractors | null | null | null | null | math.DS | http://creativecommons.org/licenses/by/4.0/ | We state necessary and sufficient conditions for the existence of $T$-discrete exponential attractors for semigroups in complete metric spaces. These conditions are formulated in terms of a covering condition for iterates of the absorbing set under the time evolution of the semigroup and imply the existence and finite-dimensionality of the global attractor. We then review, generalize and compare existing construction methods for exponential attractors and show that they all imply the covering condition. Furthermore, we relate the results and concept of $T$-discrete exponential attractors to the classical notion of exponential attractors. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:22:20 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Czaja",
"Radoslaw",
""
],
[
"Sonner",
"Stefanie",
""
]
] |
2506.10389 | Yuhang Chen | Yuhang Chen, Zhen Tan, Tianlong Chen | EQA-RM: A Generative Embodied Reward Model with Test-time Scaling | preprint | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Reward Models (RMs), vital for large model alignment, are underexplored for complex embodied tasks like Embodied Question Answering (EQA) where nuanced evaluation of agents' spatial, temporal, and logical understanding is critical yet not considered by generic approaches. We introduce EQA-RM, a novel generative multimodal reward model specifically architected for EQA, trained via our innovative Contrastive Group Relative Policy Optimization (C-GRPO) strategy to learn fine-grained behavioral distinctions. The generative nature of EQA-RM provides interpretable, structured reward feedback (beyond simple scalars), uniquely enabling test-time scaling to dynamically adjust evaluation granularity, from concise scores to detailed critiques of reasoning and grounding, at inference without retraining. Concurrently, we introduce EQARewardBench, a new benchmark built on OpenEQA for standardized EQA reward model assessment. Demonstrating high sample efficiency, EQA-RM (fine-tuning Qwen2-VL-2B-Instruct) achieves 61.9\% accuracy on EQA-RM-Bench with only 700 samples, outperforming strong proprietary baselines, including Gemini-2.5-Flash, GPT-4o, Claude-3.5-Haiku, and open-sourced state-of-the-art models such as RoVRM and VisualPRM. The code and dataset can be found here https://github.com/UNITES-Lab/EQA-RM. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:25:04 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chen",
"Yuhang",
""
],
[
"Tan",
"Zhen",
""
],
[
"Chen",
"Tianlong",
""
]
] |
2506.10390 | Yang Liu | Shicheng Yin, Kaixuan Yin, Yang Liu, Weixing Chen, Liang Lin | DART: Differentiable Dynamic Adaptive Region Tokenizer for Vision Transformer and Mamba | Code is available at https://github.com/HCPLab-SYSU/DART | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, non-convolutional models such as the Vision Transformer (ViT) and Vision Mamba (Vim) have achieved remarkable performance in computer vision tasks. However, their reliance on fixed-size patches often results in excessive encoding of background regions and omission of critical local details, especially when informative objects are sparsely distributed. To address this, we introduce a fully differentiable Dynamic Adaptive Region Tokenizer (DART), which adaptively partitions images into content-dependent patches of varying sizes. DART combines learnable region scores with piecewise differentiable quantile operations to allocate denser tokens to information-rich areas. Despite introducing only approximately 1 million (1M) additional parameters, DART improves accuracy by 2.1% on DeiT (ImageNet-1K). Unlike methods that uniformly increase token density to capture fine-grained details, DART offers a more efficient alternative, achieving 45% FLOPs reduction with superior performance. Extensive experiments on DeiT, Vim, and VideoMamba confirm that DART consistently enhances accuracy while incurring minimal or even reduced computational overhead. Code is available at https://github.com/HCPLab-SYSU/DART. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:25:37 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yin",
"Shicheng",
""
],
[
"Yin",
"Kaixuan",
""
],
[
"Liu",
"Yang",
""
],
[
"Chen",
"Weixing",
""
],
[
"Lin",
"Liang",
""
]
] |
2506.10391 | Yuanyi Song | Yuanyi Song, Pumeng Lyu, Ben Fei, Fenghua Ling, Wanli Ouyang, Lei Bai | ReconMOST: Multi-Layer Sea Temperature Reconstruction with Observations-Guided Diffusion | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate reconstruction of ocean is essential for reflecting global climate dynamics and supporting marine meteorological research. Conventional methods face challenges due to sparse data, algorithmic complexity, and high computational costs, while increasing usage of machine learning (ML) method remains limited to reconstruction problems at the sea surface and local regions, struggling with issues like cloud occlusion. To address these limitations, this paper proposes ReconMOST, a data-driven guided diffusion model framework for multi-layer sea temperature reconstruction. Specifically, we first pre-train an unconditional diffusion model using a large collection of historical numerical simulation data, enabling the model to attain physically consistent distribution patterns of ocean temperature fields. During the generation phase, sparse yet high-accuracy in-situ observational data are utilized as guidance points for the reverse diffusion process, generating accurate reconstruction results. Importantly, in regions lacking direct observational data, the physically consistent spatial distribution patterns learned during pre-training enable implicitly guided and physically plausible reconstructions. Our method extends ML-based SST reconstruction to a global, multi-layer setting, handling over 92.5% missing data while maintaining reconstruction accuracy, spatial resolution, and superior generalization capability. We pre-train our model on CMIP6 numerical simulation data and conduct guided reconstruction experiments on CMIP6 and EN4 analysis data. The results of mean squared error (MSE) values achieve 0.049 on guidance, 0.680 on reconstruction, and 0.633 on total, respectively, demonstrating the effectiveness and robustness of the proposed framework. Our source code is available at https://github.com/norsheep/ReconMOST. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:27:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Song",
"Yuanyi",
""
],
[
"Lyu",
"Pumeng",
""
],
[
"Fei",
"Ben",
""
],
[
"Ling",
"Fenghua",
""
],
[
"Ouyang",
"Wanli",
""
],
[
"Bai",
"Lei",
""
]
] |
2506.10392 | Tikaram Subedi | Dibyasman Sarma and Tikaram Subedi | The probability that the product of k elements in a finite ring is zero | null | null | null | null | math.RA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, for a fixed integer $k\ge 2$, we study the probability that the product of $k$ randomly chosen elements in a finite commutative ring $R$ is zero, which we denote by $zp_{_k}(R)$. We investigate bounds for $zp_{_k}(R)$ that turn out to be sharp bounds for certain classes of rings. Further, we determine the maximum value of $zp_{_k}(R)$ that can be obtained for any ring $R$, and classify all rings within some specific range of $zp_{_k}(R)$. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:28:06 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Sarma",
"Dibyasman",
""
],
[
"Subedi",
"Tikaram",
""
]
] |
2506.10393 | Yuang Pan | Yuang Pan, Liang Si, Miao Yang, Ning Han, Li Zhang, Qiaolu Chen, Rui Zhao, Fujia Chen, Yudong Ren, Wenhao Li, Yuze Hu, Mingyu Tong, Xinrui Li, Junyao Wu, Ronghao Bao, Weiqiu Chen, Yang Long, Bin Wu, Hongsheng Chen, Baile Zhang, Yihao Yang | Realization of Weyl elastic metamaterials with spin skyrmions | 14 pages, 4 figures | null | null | null | physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Topological elastic metamaterials provide a topologically robust way to manipulate the phononic energy and information beyond the conventional approaches. Among various topological elastic metamaterials, Weyl elastic metamaterials stand out, as they are unique to three dimensions and exhibit numerous intriguing phenomena and potential applications. To date, however, the realization of Weyl elastic metamaterials remains elusive, primarily due to the full-vectoral nature of elastic waves and the complicated couplings between polarizations, leading to complicated and tangled three-dimensional (3D) bandstructures that unfavorable for experimental demonstration. Here, we overcome the challenge and realize an ideal, 3D printed, all-metallic Weyl elastic metamaterial with low dissipation losses. Notably, the elastic spin of the excitations around the Weyl points exhibits skyrmion textures, a topologically stable structure in real space. Utilizing 3D laser vibrometry, we reveal the projection of the Weyl points, the Fermi arcs and the unique spin characteristics of the topological surface states. Our work extends the Weyl metamaterials to elastic waves and paves a topological way to robust manipulation of elastic waves in 3D space. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:31:47 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Pan",
"Yuang",
""
],
[
"Si",
"Liang",
""
],
[
"Yang",
"Miao",
""
],
[
"Han",
"Ning",
""
],
[
"Zhang",
"Li",
""
],
[
"Chen",
"Qiaolu",
""
],
[
"Zhao",
"Rui",
""
],
[
"Chen",
"Fujia",
""
],
[
"Ren",
"Yudong",
""
],
[
"Li",
"Wenhao",
""
],
[
"Hu",
"Yuze",
""
],
[
"Tong",
"Mingyu",
""
],
[
"Li",
"Xinrui",
""
],
[
"Wu",
"Junyao",
""
],
[
"Bao",
"Ronghao",
""
],
[
"Chen",
"Weiqiu",
""
],
[
"Long",
"Yang",
""
],
[
"Wu",
"Bin",
""
],
[
"Chen",
"Hongsheng",
""
],
[
"Zhang",
"Baile",
""
],
[
"Yang",
"Yihao",
""
]
] |
2506.10394 | Alice Publications | ALICE Collaboration | Study of $\langle p_{\rm T} \rangle$ and its higher moments, and extraction of the speed of sound in Pb-Pb collisions with ALICE | 35 pages, 10 captioned figures, 2 tables, authors from page 29, submitted to Journal of High Energy Physics, figures at http://alice-publications.web.cern.ch/node/12592 | null | null | CERN-EP-2025-126 | nucl-ex | http://creativecommons.org/licenses/by/4.0/ | Ultrarelativistic heavy-ion collisions create a quark-gluon plasma (QGP), a hot and dense state of strongly interacting QCD matter. In ultracentral collisions, the QGP volume remains nearly constant event-by-event, while its total entropy can fluctuate due to quantum effects, leading to temperature variations. These features allow the correlation between the mean transverse momentum $(\langle p_{\mathrm{T}} \rangle)$ of charged hadrons and their multiplicity to serve as a probe of the QGP's speed of sound, $c_{s}$. This study extracts $c_{s}$ by analyzing the relative increase in $\langle p_{\mathrm{T}} \rangle$ with respect to the charged-particle density $(\langle \mathrm{d}N_\mathrm{ch}/\mathrm{d}\eta \rangle)$ at midrapidity in Pb-Pb collisions at $\sqrt{s_{\mathrm{NN}}} = 5.02~\mathrm{TeV}$, using data from the ALICE detector. Centrality is determined with estimators based on multiplicity and transverse energy, applying a pseudorapidity gap to reduce selection biases. The extracted value of $c_{s}^{2}$ is found to strongly depend on the employed centrality estimator and ranges between $0.1146 \pm 0.0028 \,\mathrm{(stat.)} \pm 0.0065 \,\mathrm{(syst.)}$ and $0.4374 \pm 0.0006 \mathrm{(stat.)} \pm 0.0184 \mathrm{(syst.)}$ in natural units. Additionally, the event-by-event $[p_{\mathrm{T}}]$ distribution is studied through its variance, skewness, and kurtosis. A pronounced decrease in the self-normalized variance and a peak followed by a drop in skewness suggest the suppression of impact-parameter fluctuations in ultracentral collisions. These observations provide new insights into the thermodynamic properties and initial-state fluctuations of the QGP. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:34:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"ALICE Collaboration",
"",
""
]
] |
2506.10395 | Zhiyang Xu | Zhiyang Xu, Jiuhai Chen, Zhaojiang Lin, Xichen Pan, Lifu Huang, Tianyi Zhou, Madian Khabsa, Qifan Wang, Di Jin, Michihiro Yasunaga, Lili Yu, Xi Victoria Lin, Shaoliang Nie | Pisces: An Auto-regressive Foundation Model for Image Understanding and Generation | Unified image understanding and generation model | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Recent advances in large language models (LLMs) have enabled multimodal foundation models to tackle both image understanding and generation within a unified framework. Despite these gains, unified models often underperform compared to specialized models in either task. A key challenge in developing unified models lies in the inherent differences between the visual features needed for image understanding versus generation, as well as the distinct training processes required for each modality. In this work, we introduce Pisces, an auto-regressive multimodal foundation model that addresses this challenge through a novel decoupled visual encoding architecture and tailored training techniques optimized for multimodal generation. Combined with meticulous data curation, pretraining, and finetuning, Pisces achieves competitive performance in both image understanding and image generation. We evaluate Pisces on over 20 public benchmarks for image understanding, where it demonstrates strong performance across a wide range of tasks. Additionally, on GenEval, a widely adopted benchmark for image generation, Pisces exhibits robust generative capabilities. Our extensive analysis reveals the synergistic relationship between image understanding and generation, and the benefits of using separate visual encoders, advancing the field of unified multimodal models. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:37:34 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Xu",
"Zhiyang",
""
],
[
"Chen",
"Jiuhai",
""
],
[
"Lin",
"Zhaojiang",
""
],
[
"Pan",
"Xichen",
""
],
[
"Huang",
"Lifu",
""
],
[
"Zhou",
"Tianyi",
""
],
[
"Khabsa",
"Madian",
""
],
[
"Wang",
"Qifan",
""
],
[
"Jin",
"Di",
""
],
[
"Yasunaga",
"Michihiro",
""
],
[
"Yu",
"Lili",
""
],
[
"Lin",
"Xi Victoria",
""
],
[
"Nie",
"Shaoliang",
""
]
] |
2506.10396 | Kenya Tasuki | Jonathan Harper, Ali Mollabashi, Tadashi Takayanagi, Kenya Tasuki | Multi-entropy and the Dihedral Measures at Quantum Critical Points | 20 pages, 16 figures | null | null | YITP-25-87 | hep-th cond-mat.stat-mech quant-ph | http://creativecommons.org/licenses/by/4.0/ | The multi-entropy and dihedral measures are a class of tractable measures for multi-partite entanglement, which are labeled by the R\'enyi index (or replica number) $n$ as in the R\'enyi entanglement entropy. The purpose of this article is to demonstrate that these quantities are new useful probes of quantum critical points by examining concrete examples. In particular, we compute the multi-entropy and dihedral measures in the $1+1$ dimensional massless free scalar field theory on a lattice and in the transverse-field Ising model. For $n=2$, we find that the numerical results in both lattice theories quantitatively agree with those from conformal field theoretic calculations. For $n=3$ and $n=4$, we provide new predictions of these measures for the massless scalar field theory. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:37:52 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Harper",
"Jonathan",
""
],
[
"Mollabashi",
"Ali",
""
],
[
"Takayanagi",
"Tadashi",
""
],
[
"Tasuki",
"Kenya",
""
]
] |
2506.10397 | Mir Mohammad Yousuf | Mir Mohammad Yousuf, Shabir Ahmad Sofi | Bug Classification in Quantum Software: A Rule-Based Framework and Its Evaluation | 25 pages, 5 figures | null | null | null | cs.SE cs.CY cs.DC | http://creativecommons.org/licenses/by/4.0/ | Accurate classification of software bugs is essential for improving software quality. This paper presents a rule-based automated framework for classifying issues in quantum software repositories by bug type, category, severity, and impacted quality attributes, with additional focus on quantum-specific bug types. The framework applies keyword and heuristic-based techniques tailored to quantum computing. To assess its reliability, we manually classified a stratified sample of 4,984 issues from a dataset of 12,910 issues across 36 Qiskit repositories. Automated classifications were compared with ground truth using accuracy, precision, recall, and F1-score. The framework achieved up to 85.21% accuracy, with F1-scores ranging from 0.7075 (severity) to 0.8393 (quality attribute). Statistical validation via paired t-tests and Cohen's Kappa showed substantial to almost perfect agreement for bug type (k = 0.696), category (k = 0.826), quality attribute (k = 0.818), and quantum-specific bug type (k = 0.712). Severity classification showed slight agreement (k = 0.162), suggesting room for improvement. Large-scale analysis revealed that classical bugs dominate (67.2%), with quantum-specific bugs at 27.3%. Frequent bug categories included compatibility, functional, and quantum-specific defects, while usability, maintainability, and interoperability were the most impacted quality attributes. Most issues (93.7%) were low severity; only 4.3% were critical. A detailed review of 1,550 quantum-specific bugs showed that over half involved quantum circuit-level problems, followed by gate errors and hardware-related issues. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:42:10 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yousuf",
"Mir Mohammad",
""
],
[
"Sofi",
"Shabir Ahmad",
""
]
] |
2506.10398 | Yu-Xiao Liu | Hong-Yue Jiang and Yu-Xiao Liu | Complexity equals anything for multi-horizon black holes | 29 pages, 20 figures | null | null | null | hep-th gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the ``complexity equals anything" proposal with codimension-one and codimension-zero gravitational observables for multi-horizon black holes, using the Bardeen-AdS class black hole as an example. In particular, we compare the results with the ``complexity equals volume" (CV) proposal and find that the ``complexity equals anything" ($\mathcal{C}_{\text{gen}}$) enables the probing of a more complete black hole interior, that is, all spacetime regions where the blackening factor $f(r)<0$. This is the advantage brought by the flexibility of this holographic complexity conjecture. In addition, we compute the complexity derived from various geometric quantities and show that these constructions can effectively differentiate the distinct interior regions of the black hole. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:45:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Jiang",
"Hong-Yue",
""
],
[
"Liu",
"Yu-Xiao",
""
]
] |
2506.10399 | Zhaoxuan Kan | Zhaoxuan Kan, Husheng Han, Shangyi Shi, Tenghui Hua, Hang Lu, Xiaowei Li, Jianan Mu, Xing Hu | FicGCN: Unveiling the Homomorphic Encryption Efficiency from Irregular Graph Convolutional Networks | Accepted by ICML 2025 | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Graph Convolutional Neural Networks (GCNs) have gained widespread popularity in various fields like personal healthcare and financial systems, due to their remarkable performance. Despite the growing demand for cloud-based GCN services, privacy concerns over sensitive graph data remain significant. Homomorphic Encryption (HE) facilitates Privacy-Preserving Machine Learning (PPML) by allowing computations to be performed on encrypted data. However, HE introduces substantial computational overhead, particularly for GCN operations that require rotations and multiplications in matrix products. The sparsity of GCNs offers significant performance potential, but their irregularity introduces additional operations that reduce practical gains. In this paper, we propose FicGCN, a HE-based framework specifically designed to harness the sparse characteristics of GCNs and strike a globally optimal balance between aggregation and combination operations. FicGCN employs a latency-aware packing scheme, a Sparse Intra-Ciphertext Aggregation (SpIntra-CA) method to minimize rotation overhead, and a region-based data reordering driven by local adjacency structure. We evaluated FicGCN on several popular datasets, and the results show that FicGCN achieved the best performance across all tested datasets, with up to a 4.10x improvement over the latest design. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:46:07 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Kan",
"Zhaoxuan",
""
],
[
"Han",
"Husheng",
""
],
[
"Shi",
"Shangyi",
""
],
[
"Hua",
"Tenghui",
""
],
[
"Lu",
"Hang",
""
],
[
"Li",
"Xiaowei",
""
],
[
"Mu",
"Jianan",
""
],
[
"Hu",
"Xing",
""
]
] |
2506.10400 | Timur Sahin | Gizay Yolalan, Timur \c{S}ahin, Sel\c{c}uk Bilir, Olcay Plevne | Stellar Ancestry Unlocked: Chemical and Orbital Clues Link Metal Poor Stars to Globular Clusters and the Galaxy's Thick Disk | 35 pages, including 18 figures,10 tables, accepted for publication in the Astronomical Journal | null | null | null | astro-ph.GA astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | This study presents a detailed chemical, kinematic, and orbital dynamic analysis of five metal poor stars in the solar neighborhood: HD 2665, HD 5916, HD 122956, HD 189349, and HD 218857. Using high-resolution spectroscopic data from the ELODIE and ESPaDOnS instruments, we derived elemental abundances for 29 species (25 elements: C, O, Na, Mg, Al, Si, S, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Sr, Y, Zr, Ba, Ce, Nd, and Sm) via LTE based analysis with ATLAS9 model atmospheres. Notably, we report first time detections of Ce and Nd in HD2665; Al, V, Sm, and Mn in HD5916; Al in HD 122956; and C, O, S, Sc, Mn, Co, Cu, Zn, Sr, Zr, Nd, and Sm in HD 189349. Dynamical and chemical diagnostics reveal distinct origins: HD 2665 shows strong orbital and chemical similarity to GC NGC 5139 ($\omega$ Cen), while HD 218857 exhibits chemodynamic signatures consistent with NGC 5634. HD 122956 aligns with NGC 6864 (M75), though intriguingly shares age, metallicity ([Fe/H]), and [Mg/Fe] ratios with NGC 6517, a cluster whose reported abundances are derived solely from one star with APOGEE $H$-band spectroscopic measurements, as no optical spectroscopic data exist for its members. In contrast, HD 5916 and HD 189349 exhibit kinematic and chemical properties consistent with the field-star population and are classified as thick disk members. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:47:43 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yolalan",
"Gizay",
""
],
[
"Şahin",
"Timur",
""
],
[
"Bilir",
"Selçuk",
""
],
[
"Plevne",
"Olcay",
""
]
] |
2506.10401 | Jiaqi Lv | Jiaqi Lv, Xufeng He, Yanchen Liu, Xu Dai, Yang Hu, Shouyi Yin | HPCTransCompile: An AI Compiler Generated Dataset for High-Performance CUDA Transpilation and LLM Preliminary Exploration | null | null | null | null | cs.DC cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid growth of deep learning has driven exponential increases in model parameters and computational demands. NVIDIA GPUs and their CUDA-based software ecosystem provide robust support for parallel computing, significantly alleviating computational bottlenecks. Meanwhile, due to the cultivation of user programming habits and the high performance of GPUs, the CUDA ecosystem has established a dominant position in the field of parallel software. This dominance requires other hardware platforms to support CUDA-based software with performance portability. However, translating CUDA code to other platforms poses significant challenges due to differences in parallel programming paradigms and hardware architectures. Existing approaches rely on language extensions, domain-specific languages (DSLs), or compilers but face limitations in workload coverage and generalizability. Moreover, these methods often incur substantial development costs. Recently, LLMs have demonstrated extraordinary potential in various vertical domains, especially in code-related tasks. However, the performance of existing LLMs in CUDA transpilation, particularly for high-performance code, remains suboptimal. The main reason for this limitation lies in the lack of high-quality training datasets. To address these challenges, we propose a novel framework for generating high-performance CUDA and corresponding platform code pairs, leveraging AI compiler and automatic optimization technology. We further enhance the framework with a graph-based data augmentation method and introduce HPCTransEval, a benchmark for evaluating LLM performance on CUDA transpilation. We conduct experiments using CUDA-to-CPU transpilation as a case study on leading LLMs. The result demonstrates that our framework significantly improves CUDA transpilation, highlighting the potential of LLMs to address compatibility challenges within the CUDA ecosystem. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:48:33 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Lv",
"Jiaqi",
""
],
[
"He",
"Xufeng",
""
],
[
"Liu",
"Yanchen",
""
],
[
"Dai",
"Xu",
""
],
[
"Hu",
"Yang",
""
],
[
"Yin",
"Shouyi",
""
]
] |
2506.10402 | Marios Zacharias | Marios Zacharias and George Volonakis and Laurent Pedesseau and Claudine Katan and Feliciano Giustino and Jacky Even | Roadmap for electronic structure, anharmonicity, and electron-phonon calculations in locally disordered inorganic and hybrid halide perovskites | null | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The role of data in modern materials science becomes more valuable and accurate when effects such as electron-phonon coupling and anharmonicity are included, providing a more realistic representation of finite-temperature material behavior. Furthermore, positional polymorphism, characterized by correlated local atomic disorder usually not reported by standard diffraction techniques, is a critical yet underexplored factor in understanding the electronic structure and transport properties of energy-efficient materials, like halide perovskites. In this manuscript, we present a first-principles methodology for locally disordered (polymorphous) cubic inorganic and hybrid halide perovskites, rooted in the special displacement method, that offers a systematic and alternative approach to molecular dynamics for exploring finite-temperature properties. By enabling a unified and efficient treatment of anharmonic lattice dynamics, electron-phonon coupling, and positional polymorphism, our approach generates essential data to predict temperature-dependent phonon properties, free energies, band gaps, and effective masses. Designed with a high-throughput spirit, this framework has been applied across a range of inorganic and hybrid halide perovskites: CsPbI3, CsPbBr3, CsSnI3, CsPbCl3, MAPbI3, MAPbBr3, MASnI3, MAPbCl3, FAPbI3, FAPbBr3, FASnI3, and FAPbCl3. We provide a comprehensive comparison between theoretical and experimental results and we systematically uncover trends and insights into their electronic and thermal behavior. For all compounds, we demonstrate strong and consistent correlations between local structural disorder, band gap openings, and effective mass enhancements. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:52:20 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zacharias",
"Marios",
""
],
[
"Volonakis",
"George",
""
],
[
"Pedesseau",
"Laurent",
""
],
[
"Katan",
"Claudine",
""
],
[
"Giustino",
"Feliciano",
""
],
[
"Even",
"Jacky",
""
]
] |
2506.10403 | Tzu-Heng Huang | Tzu-Heng Huang, Harit Vishwakarma, Frederic Sala | Time To Impeach LLM-as-a-Judge: Programs are the Future of Evaluation | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) are widely used to evaluate the quality of LLM generations and responses, but this leads to significant challenges: high API costs, uncertain reliability, inflexible pipelines, and inherent biases. To address these, we introduce PAJAMA (Program-As-a-Judge for Automated Model Assessment), a new alternative that uses LLMs to synthesize executable judging programs instead of directly scoring responses. These synthesized programs can be stored and run locally, costing orders of magnitude less while providing interpretable, and auditable judging logic that can be easily adapted. Program-based judges mitigate biases, improving judgment consistency by 15.83% and reducing biased responses by 23.7% on average compared to a Qwen2.5-14B-based LLM-as-a-judge. When program judgments are distilled into a model, PAJAMA outperforms LLM-as-a-judge on the challenging CHAT-HARD subset of RewardBench, outperforming metrics by 2.19% on Prometheus and 8.67% on the JudgeLM dataset, all at three orders of magnitude lower cost. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:53:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Huang",
"Tzu-Heng",
""
],
[
"Vishwakarma",
"Harit",
""
],
[
"Sala",
"Frederic",
""
]
] |
2506.10404 | Bryan Shaddy | Bryan Shaddy, Brianna Binder, Agnimitra Dasgupta, Haitong Qin, James Haley, Angel Farguell, Kyle Hilburn, Derek V. Mallia, Adam Kochanski, Jan Mandel, and Assad Oberai | Generative Algorithms for Wildfire Progression Reconstruction from Multi-Modal Satellite Active Fire Measurements and Terrain Height | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Increasing wildfire occurrence has spurred growing interest in wildfire spread prediction. However, even the most complex wildfire models diverge from observed progression during multi-day simulations, motivating need for data assimilation. A useful approach to assimilating measurement data into complex coupled atmosphere-wildfire models is to estimate wildfire progression from measurements and use this progression to develop a matching atmospheric state. In this study, an approach is developed for estimating fire progression from VIIRS active fire measurements, GOES-derived ignition times, and terrain height data. A conditional Generative Adversarial Network is trained with simulations of historic wildfires from the atmosphere-wildfire model WRF-SFIRE, thus allowing incorporation of WRF-SFIRE physics into estimates. Fire progression is succinctly represented by fire arrival time, and measurements for training are obtained by applying an approximate observation operator to WRF-SFIRE solutions, eliminating need for satellite data during training. The model is trained on tuples of fire arrival times, measurements, and terrain, and once trained leverages measurements of real fires and corresponding terrain data to generate samples of fire arrival times. The approach is validated on five Pacific US wildfires, with results compared against high-resolution perimeters measured via aircraft, finding an average Sorensen-Dice coefficient of 0.81. The influence of terrain height on the arrival time inference is also evaluated and it is observed that terrain has minimal influence when the inference is conditioned on satellite measurements. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:54:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Shaddy",
"Bryan",
""
],
[
"Binder",
"Brianna",
""
],
[
"Dasgupta",
"Agnimitra",
""
],
[
"Qin",
"Haitong",
""
],
[
"Haley",
"James",
""
],
[
"Farguell",
"Angel",
""
],
[
"Hilburn",
"Kyle",
""
],
[
"Mallia",
"Derek V.",
""
],
[
"Kochanski",
"Adam",
""
],
[
"Mandel",
"Jan",
""
],
[
"Oberai",
"Assad",
""
]
] |
2506.10405 | Antonin Novak | Ond\v{r}ej Benedikt, Istv\'an M\'odos, Antonin Novak, Zden\v{e}k Hanz\'alek | Green Scheduling with Time-of-Use Tariffs and Machine States: Optimizing Energy Cost via Branch-and-Bound and Bin Packing Strategies | null | null | null | null | math.OC | http://creativecommons.org/licenses/by/4.0/ | This paper presents a branch-and-bound algorithm, enhanced with bin packing strategies, for scheduling under variable energy pricing and power-saving states. The proposed algorithm addresses the 1,TOU|states|TEC problem, which involves scheduling jobs to minimize total energy cost (TEC) while considering time-of-use (TOU) electricity prices and different machine states (e.g., processing, idle, off). Key innovations include instance pre-processing for rapid lower bound calculations, a novel branching scheme combined with initializations, a block-finding primal heuristic, and a tighter lower bound for jobs with non-coprime processing times. These enhancements result in an efficient algorithm capable of solving benchmark instances with real energy prices with 200 jobs more than 100 times faster than existing state-of-the-art methods. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:55:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Benedikt",
"Ondřej",
""
],
[
"Módos",
"István",
""
],
[
"Novak",
"Antonin",
""
],
[
"Hanzálek",
"Zdeněk",
""
]
] |
2506.10406 | Yuhua Jiang | Yuhua Jiang, Yuwen Xiong, Yufeng Yuan, Chao Xin, Wenyuan Xu, Yu Yue, Qianchuan Zhao, Lin Yan | PAG: Multi-Turn Reinforced LLM Self-Correction with Policy as Generative Verifier | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have demonstrated impressive capabilities in complex reasoning tasks, yet they still struggle to reliably verify the correctness of their own outputs. Existing solutions to this verification challenge often depend on separate verifier models or require multi-stage self-correction training pipelines, which limit scalability. In this paper, we propose Policy as Generative Verifier (PAG), a simple and effective framework that empowers LLMs to self-correct by alternating between policy and verifier roles within a unified multi-turn reinforcement learning (RL) paradigm. Distinct from prior approaches that always generate a second attempt regardless of model confidence, PAG introduces a selective revision mechanism: the model revises its answer only when its own generative verification step detects an error. This verify-then-revise workflow not only alleviates model collapse but also jointly enhances both reasoning and verification abilities. Extensive experiments across diverse reasoning benchmarks highlight PAG's dual advancements: as a policy, it enhances direct generation and self-correction accuracy; as a verifier, its self-verification outperforms self-consistency. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 06:59:35 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Jiang",
"Yuhua",
""
],
[
"Xiong",
"Yuwen",
""
],
[
"Yuan",
"Yufeng",
""
],
[
"Xin",
"Chao",
""
],
[
"Xu",
"Wenyuan",
""
],
[
"Yue",
"Yu",
""
],
[
"Zhao",
"Qianchuan",
""
],
[
"Yan",
"Lin",
""
]
] |
2506.10407 | Daizhan Cheng Dr | Daizhan Cheng | Semi-Tensor-Product Based Convolutional Neural Networks | null | null | null | null | eess.SY cs.AI cs.CV cs.SY | http://creativecommons.org/licenses/by/4.0/ | The semi-tensor product (STP) of vectors is a generalization of conventional inner product of vectors, which allows the factor vectors to of different dimensions. This paper proposes a domain-based convolutional product (CP). Combining domain-based CP with STP of vectors, a new CP is proposed. Since there is no zero or any other padding, it can avoid the junk information caused by padding. Using it, the STP-based convolutional neural network (CNN) is developed. Its application to image and third order signal identifications is considered. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:00:19 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Cheng",
"Daizhan",
""
]
] |
2506.10408 | Ziyue Li Dr | Jintao Liang, Gang Su, Huifeng Lin, You Wu, Rui Zhao, Ziyue Li | Reasoning RAG via System 1 or System 2: A Survey on Reasoning Agentic Retrieval-Augmented Generation for Industry Challenges | null | null | null | null | cs.AI cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Retrieval-Augmented Generation (RAG) has emerged as a powerful framework to overcome the knowledge limitations of Large Language Models (LLMs) by integrating external retrieval with language generation. While early RAG systems based on static pipelines have shown effectiveness in well-structured tasks, they struggle in real-world scenarios requiring complex reasoning, dynamic retrieval, and multi-modal integration. To address these challenges, the field has shifted toward Reasoning Agentic RAG, a paradigm that embeds decision-making and adaptive tool use directly into the retrieval process. In this paper, we present a comprehensive review of Reasoning Agentic RAG methods, categorizing them into two primary systems: predefined reasoning, which follows fixed modular pipelines to boost reasoning, and agentic reasoning, where the model autonomously orchestrates tool interaction during inference. We analyze representative techniques under both paradigms, covering architectural design, reasoning strategies, and tool coordination. Finally, we discuss key research challenges and propose future directions to advance the flexibility, robustness, and applicability of reasoning agentic RAG systems. Our collection of the relevant research has been organized into a https://github.com/ByebyeMonica/Reasoning-Agentic-RAG. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:01:56 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Liang",
"Jintao",
""
],
[
"Su",
"Gang",
""
],
[
"Lin",
"Huifeng",
""
],
[
"Wu",
"You",
""
],
[
"Zhao",
"Rui",
""
],
[
"Li",
"Ziyue",
""
]
] |
2506.10409 | Bing Sun | Hao Wang, Zhi Xiao, Bing Sun | Analytical solutions of CPT-odd Maxwell equations in Schwarzschild spacetime | Accepted for publication in Chinese Physics C. 14 pages | null | null | null | gr-qc hep-ph | http://creativecommons.org/licenses/by/4.0/ | In this work, we present the CPT-violating (CPTV) Maxwell equations in curved spacetime using the Newman-Penrose (NP) formalism. We obtain a semi-analytical solution to the Maxwell equations in Schwarzschild spacetime under the assumption that the CPT-odd $\left(k_{AF}\right)^\mu$ term exhibits spherical symmetry in the Schwarzschild background. Retaining only terms up to linear order in the $\left(k_{AF}\right)^\mu$ coefficient, we obtain perturbative solutions by treating the solutions of the Lorentz-invariant Maxwell equations as the zeroth-order approximation and incorporating the $\left(k_{AF}\right)^\mu$ terms as an additional source term alongside the external charge current. Each resulting NP scalar field can be factorized into two components: the radial component is expressed in terms of hypergeometric functions, while the angular component is described by spin-weighted spherical harmonics. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:05:12 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Wang",
"Hao",
""
],
[
"Xiao",
"Zhi",
""
],
[
"Sun",
"Bing",
""
]
] |
2506.10410 | Tobias Dornheim | Panagiotis Tolias and Jan Vorberger and Tobias Dornheim | Exact series expansion for even frequency moments of the dynamic structure factor | null | null | null | null | physics.plasm-ph physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An exact series representation of the even frequency moments of the dynamic structure factor is derived. Truncations are proposed that allow to evaluate the explicitly unknown second, fourth and fifth frequency moments for the finite temperature uniform electron gas. Their applicability range in terms of degeneracy parameter and wavenumber is determined by exploiting the non-interacting limit and by comparing with the quasi-exact results of path integral Monte Carlo simulations. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:05:16 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Tolias",
"Panagiotis",
""
],
[
"Vorberger",
"Jan",
""
],
[
"Dornheim",
"Tobias",
""
]
] |
2506.10411 | Matthias Hoelzl | S. Ratynskaia, M. Hoelzl, E. Nardon, P. Aleynikov, F.J. Artola, V. Bandaru, M. Beidler, B. Breizman, D. del-Castillo-Negrete, M. De Angeli, V. Dimitriou, R. Ding, J. Eriksson, O. Ficker, R.S. Granetz, E. Hollmann, M. Hoppe, M. Houry, I. Jepu, H.R. Koslowski, C. Liu, J.R. Martin-Solis, G. Pautasso, Y. Peneliau, R.A. Pitts, G.I. Pokol, C. Reux, U. Sheikh, S.A. Silburn, T. Tang, R.A. Tinguely, P. Tolias, E. Tomesova, R. Villari | Runaway electron-induced plasma facing component damage in tokamaks | Submitted for publication in the journal Plasma Physics and Controlled Fusion | null | null | null | physics.plasm-ph physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This Roadmap article addresses the critical and multifaceted challenge of plasma-facing component (PFC) damage caused by runaway electrons (REs) in tokamaks, a phenomenon that poses a significant threat to the viability and longevity of future fusion reactors such as ITER and DEMO. The dramatically increased RE production expected in future high-current tokamaks makes it difficult to avoid or mitigate REs when a plasma discharge terminates abnormally. Preventing damage from the intense localised heat loads REs can cause requires a holistic approach that considers plasma, REs and PFC damage. Despite decades of progress in understanding the physics of REs and the thermomechanical response of PFCs, their complex interplay remains poorly understood. This document aims to initiate a coordinated, interdisciplinary approach to bridge this gap by reviewing experimental evidence, advancing diagnostic capabilities, and improving modelling tools across different scales, dimensionalities and fidelities. Key topics include RE beam formation and transport, damage mechanisms in brittle and metallic PFCs, and observations in major facilities such as JET, DIII-D, WEST and EAST. The Roadmap emphasises the urgency of predictive, high-fidelity modelling validated against well-diagnosed controlled experiments, particularly in the light of recent changes in ITER's wall material strategy and the growing importance of private sector initiatives. Each section of the article is written to provide a concise overview of one area of this multidisciplinary subject, with an assessment of the status, a look at current and future challenges, and a brief summary. The ultimate goal of this initiative is to guide future mitigation strategies and design resilient components that can withstand the loads imposed by REs, thus ensuring the safe and sustainable operation of the next generation of fusion power plants. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:06:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ratynskaia",
"S.",
""
],
[
"Hoelzl",
"M.",
""
],
[
"Nardon",
"E.",
""
],
[
"Aleynikov",
"P.",
""
],
[
"Artola",
"F. J.",
""
],
[
"Bandaru",
"V.",
""
],
[
"Beidler",
"M.",
""
],
[
"Breizman",
"B.",
""
],
[
"del-Castillo-Negrete",
"D.",
""
],
[
"De Angeli",
"M.",
""
],
[
"Dimitriou",
"V.",
""
],
[
"Ding",
"R.",
""
],
[
"Eriksson",
"J.",
""
],
[
"Ficker",
"O.",
""
],
[
"Granetz",
"R. S.",
""
],
[
"Hollmann",
"E.",
""
],
[
"Hoppe",
"M.",
""
],
[
"Houry",
"M.",
""
],
[
"Jepu",
"I.",
""
],
[
"Koslowski",
"H. R.",
""
],
[
"Liu",
"C.",
""
],
[
"Martin-Solis",
"J. R.",
""
],
[
"Pautasso",
"G.",
""
],
[
"Peneliau",
"Y.",
""
],
[
"Pitts",
"R. A.",
""
],
[
"Pokol",
"G. I.",
""
],
[
"Reux",
"C.",
""
],
[
"Sheikh",
"U.",
""
],
[
"Silburn",
"S. A.",
""
],
[
"Tang",
"T.",
""
],
[
"Tinguely",
"R. A.",
""
],
[
"Tolias",
"P.",
""
],
[
"Tomesova",
"E.",
""
],
[
"Villari",
"R.",
""
]
] |
2506.10412 | Ching Chang | Ching Chang, Jeehyun Hwang, Yidan Shi, Haixin Wang, Wen-Chih Peng, Tien-Fu Chen, Wei Wang | Time-IMM: A Dataset and Benchmark for Irregular Multimodal Multivariate Time Series | This paper is currently under review | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Time series data in real-world applications such as healthcare, climate modeling, and finance are often irregular, multimodal, and messy, with varying sampling rates, asynchronous modalities, and pervasive missingness. However, existing benchmarks typically assume clean, regularly sampled, unimodal data, creating a significant gap between research and real-world deployment. We introduce Time-IMM, a dataset specifically designed to capture cause-driven irregularity in multimodal multivariate time series. Time-IMM represents nine distinct types of time series irregularity, categorized into trigger-based, constraint-based, and artifact-based mechanisms. Complementing the dataset, we introduce IMM-TSF, a benchmark library for forecasting on irregular multimodal time series, enabling asynchronous integration and realistic evaluation. IMM-TSF includes specialized fusion modules, including a timestamp-to-text fusion module and a multimodality fusion module, which support both recency-aware averaging and attention-based integration strategies. Empirical results demonstrate that explicitly modeling multimodality on irregular time series data leads to substantial gains in forecasting performance. Time-IMM and IMM-TSF provide a foundation for advancing time series analysis under real-world conditions. The dataset is publicly available at https://www.kaggle.com/datasets/blacksnail789521/time-imm/data, and the benchmark library can be accessed at https://anonymous.4open.science/r/IMMTSF_NeurIPS2025. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:07:22 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Chang",
"Ching",
""
],
[
"Hwang",
"Jeehyun",
""
],
[
"Shi",
"Yidan",
""
],
[
"Wang",
"Haixin",
""
],
[
"Peng",
"Wen-Chih",
""
],
[
"Chen",
"Tien-Fu",
""
],
[
"Wang",
"Wei",
""
]
] |
2506.10413 | Roopkatha Banerjee | Roopkatha Banerjee, Tejus Chandrashekar, Ananth Eswar and Yogesh Simmhan | Federated Learning within Global Energy Budget over Heterogeneous Edge Accelerators | Preprint of paper to appear in the proceedings of the 31st International European Conference on Parallel and Distributed Computing (EuroPar) | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated Learning (FL) enables collaborative model training across distributed clients while preserving data privacy. However, optimizing both energy efficiency and model accuracy remains a challenge, given device and data heterogeneity. Further, sustainable AI through a global energy budget for FL has not been explored. We propose a novel optimization problem for client selection in FL that maximizes the model accuracy within an overall energy limit and reduces training time. We solve this with a unique bi-level ILP formulation that leverages approximate Shapley values and energy-time prediction models to efficiently solve this. Our FedJoule framework achieves superior training accuracies compared to SOTA and simple baselines for diverse energy budgets, non-IID distributions, and realistic experiment configurations, performing 15% and 48% better on accuracy and time, respectively. The results highlight the effectiveness of our method in achieving a viable trade-off between energy usage and performance in FL environments. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:08:30 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Banerjee",
"Roopkatha",
""
],
[
"Chandrashekar",
"Tejus",
""
],
[
"Eswar",
"Ananth",
""
],
[
"Simmhan",
"Yogesh",
""
]
] |
2506.10414 | Ajay Sharma | Kanav Ganguly, Ajay Kumar Sharma, and Murli Manohar Verma | Variation in the size of the Photon Sphere and Black Hole Shadow in the Modified Gravity | 8 pages, 2 figures | null | null | null | gr-qc astro-ph.GA | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Black hole shadows are a widespread topic in astrophysics. This paper searches for an optical view of the black hole and the relationship between black hole shadow and photon sphere with curvature. We were inspired by the observations of Sagittarius $A^*$ and supermassive black hole M87$*$ through the Event Horizon Telescope. We have found the signature of the modified theory of gravity on the photon sphere and shadow. We considered $f(R)$ modified theory of gravity and new scalar degree of freedom $F$ appears in the expressions of photon sphere and shadow. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:09:56 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Ganguly",
"Kanav",
""
],
[
"Sharma",
"Ajay Kumar",
""
],
[
"Verma",
"Murli Manohar",
""
]
] |
2506.10415 | Yingjin Song | Yingjin Song, Yupei Du, Denis Paperno and Albert Gatt | Burn After Reading: Do Multimodal Large Language Models Truly Capture Order of Events in Image Sequences? | 27 pages, 14 figures. Accepted to ACL 2025 | null | null | null | cs.CL cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces the TempVS benchmark, which focuses on temporal grounding and reasoning capabilities of Multimodal Large Language Models (MLLMs) in image sequences. TempVS consists of three main tests (i.e., event relation inference, sentence ordering and image ordering), each accompanied with a basic grounding test. TempVS requires MLLMs to rely on both visual and linguistic modalities to understand the temporal order of events. We evaluate 38 state-of-the-art MLLMs, demonstrating that models struggle to solve TempVS, with a substantial performance gap compared to human capabilities. We also provide fine-grained insights that suggest promising directions for future research. Our TempVS benchmark data and code are available at https://github.com/yjsong22/TempVS. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:12:31 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Song",
"Yingjin",
""
],
[
"Du",
"Yupei",
""
],
[
"Paperno",
"Denis",
""
],
[
"Gatt",
"Albert",
""
]
] |
2506.10416 | Ali Vos | Ali Vosoughi, Jing Bi, Pinxin Liu, Yunlong Tang, Chenliang Xu | Can Sound Replace Vision in LLaVA With Token Substitution? | 29 pages including references and appendices | null | null | null | cs.MM cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | While multimodal systems have achieved impressive advances, they typically rely on text-aligned representations rather than directly integrating audio and visual inputs. This reliance can limit the use of acoustic information in tasks requiring nuanced audio understanding. In response, SoundCLIP explores direct audio-visual integration within multimodal large language models (MLLMs) by substituting CLIP's visual tokens with audio representations and selecting sound-relevant patch tokens in models such as LLaVA. We investigate two configurations: (1) projecting audio features into CLIP's visual manifold via a multilayer perceptron trained with InfoNCE on paired audio-video segments, and (2) preserving raw audio embeddings with minimal dimensional adjustments. Experiments with five state-of-the-art audio encoders reveal a fundamental trade-off. While audio-to-video retrieval performance increases dramatically (up to 44 percentage points in Top-1 accuracy) when audio is projected into CLIP's space, text generation quality declines. Encoders pre-trained with text supervision (CLAP, Whisper, ImageBind) maintain stronger generative capabilities than those focused primarily on audiovisual alignment (Wav2CLIP, AudioCLIP), highlighting the value of language exposure for generation tasks. We introduce WhisperCLIP, an architecture that fuses intermediate representations from Whisper, as well as AudioVisual Event Evaluation (AVE-2), a dataset of 580,147 three-second audiovisual clips with fine-grained alignment annotations. Our findings challenge the assumption that stronger cross-modal alignment necessarily benefits all multimodal tasks; instead, a Pareto frontier emerges wherein optimal performance depends on balancing retrieval accuracy with text generation quality. Codes and datasets: https://github.com/ali-vosoughi/SoundCLIP. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:14:23 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Vosoughi",
"Ali",
""
],
[
"Bi",
"Jing",
""
],
[
"Liu",
"Pinxin",
""
],
[
"Tang",
"Yunlong",
""
],
[
"Xu",
"Chenliang",
""
]
] |
2506.10417 | Darius Modirrousta-Galian | Darius Modirrousta-Galian and Jun Korenaga | On the Efficacy of Ocean Formation with a Primordial Hydrogen Atmosphere | Accepted for publication in the Journal of Geophysical Research - Planets | null | null | null | astro-ph.EP physics.ao-ph physics.geo-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It has been suggested that Earth's present water budget formed from oxidation reactions between its initial hydrogen-rich primordial atmosphere and its magma ocean. Here we examine this hypothesis by building a comprehensive atmosphere-magma ocean model. We find that water formation is unlikely for two reasons. First, any water formed from oxidation reactions in the magma ocean would quickly outgass because of the water-poor atmosphere above. Second, the top boundary layer of the magma ocean becomes stable against convection because the oxidation reactions produce metallic iron, which sinks to the core of a growing Earth. This iron loss makes the top boundary layer significantly more buoyant than the rest of the magma, thus becoming stable against mixing. Our results suggest that hydrogen dissolution is unlikely to play a major role in the formation of Earth's oceans. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:14:27 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Modirrousta-Galian",
"Darius",
""
],
[
"Korenaga",
"Jun",
""
]
] |
2506.10418 | Gibaek Kim | Gibaek Kim, and Jungho Kim | Efficient nanophotonic devices optimization using deep neural network trained with physics-based transfer learning (PBTL) methodology | null | null | null | null | physics.optics physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a neural network(NN)-based surrogate modeling framework for photonic device optimization, especially in domains with imbalanced feature importance and high data generation costs. Our framework, which comprises physics-based transfer learning (PBTL)-enhanced surrogate modeling and scalarized multi-objective genetic algorithms (GAs), offers a generalizable solution for photonic design automation with minimal data resources.To validate the framework, we optimize mid-infrared quantum cascade laser (QCL) structures consisting of two regions, active and injection, which have different levels of feature importance. The optimization targets include five key QCL performance metrics such as modal gain, emission wavelength, linewidth, and effective injection, extraction energies. To address the challenge of multiple local optima in the output latent space, we integrate a deep neural network total predictor (DNN-TP) with a GA, enabling scalable and nature-inspired optimization. By replacing computationally expensive numerical simulations with the DNN-TP surrogate model, the optimization achieves a speed-up of over 80,000 times, allowing large-scale exploration of the QCL design space.To improve model generalization with limited data, we introduce PBTL, which transfers knowledge from a DNN core predictor (DNN-CP) trained on active-region structures. This approach yields a 0.69 percentage increase in prediction accuracy, equivalent to a 50 percentage reduction in training data requirements, and leads to generate more feasible device structure with 60 percentage improvement in evaluation metric during optimization. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:14:30 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Kim",
"Gibaek",
""
],
[
"Kim",
"Jungho",
""
]
] |
2506.10419 | Weiying Zhao | Weiying Zhao, Aleksei Unagaev, Natalia Efremova | Data-Driven Soil Organic Carbon Sampling: Integrating Spectral Clustering with Conditioned Latin Hypercube Optimization | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Soil organic carbon (SOC) monitoring often relies on selecting representative field sampling locations based on environmental covariates. We propose a novel hybrid methodology that integrates spectral clustering - an unsupervised machine learning technique with conditioned Latin hypercube sampling (cLHS) to enhance the representativeness of SOC sampling. In our approach, spectral clustering partitions the study area into $K$ homogeneous zones using multivariate covariate data, and cLHS is then applied within each zone to select sampling locations that collectively capture the full diversity of environmental conditions. This hybrid spectral-cLHS method ensures that even minor but important environmental clusters are sampled, addressing a key limitation of vanilla cLHS which can overlook such areas. We demonstrate on a real SOC mapping dataset that spectral-cLHS provides more uniform coverage of covariate feature space and spatial heterogeneity than standard cLHS. This improved sampling design has the potential to yield more accurate SOC predictions by providing better-balanced training data for machine learning models. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:18:05 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhao",
"Weiying",
""
],
[
"Unagaev",
"Aleksei",
""
],
[
"Efremova",
"Natalia",
""
]
] |
2506.10420 | Boris Sedlak | Boris Sedlak, Alireza Furutanpey, Zihang Wang, V\'ictor Casamayor Pujol, Schahram Dustdar | Multi-dimensional Autoscaling of Processing Services: A Comparison of Agent-based Methods | null | null | null | null | cs.AI cs.DC cs.ET cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Edge computing breaks with traditional autoscaling due to strict resource constraints, thus, motivating more flexible scaling behaviors using multiple elasticity dimensions. This work introduces an agent-based autoscaling framework that dynamically adjusts both hardware resources and internal service configurations to maximize requirements fulfillment in constrained environments. We compare four types of scaling agents: Active Inference, Deep Q Network, Analysis of Structural Knowledge, and Deep Active Inference, using two real-world processing services running in parallel: YOLOv8 for visual recognition and OpenCV for QR code detection. Results show all agents achieve acceptable SLO performance with varying convergence patterns. While the Deep Q Network benefits from pre-training, the structural analysis converges quickly, and the deep active inference agent combines theoretical foundations with practical scalability advantages. Our findings provide evidence for the viability of multi-dimensional agent-based autoscaling for edge environments and encourage future work in this research direction. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:20:26 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Sedlak",
"Boris",
""
],
[
"Furutanpey",
"Alireza",
""
],
[
"Wang",
"Zihang",
""
],
[
"Pujol",
"Víctor Casamayor",
""
],
[
"Dustdar",
"Schahram",
""
]
] |
2506.10421 | Arnav Arora | Avneet Kaur and Arnav Arora | Beyond the Battlefield: Framing Analysis of Media Coverage in Conflict Reporting | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Framing used by news media, especially in times of conflict, can have substantial impact on readers' opinion, potentially aggravating the conflict itself. Current studies on the topic of conflict framing have limited insights due to their qualitative nature or only look at surface level generic frames without going deeper. In this work, we identify indicators of war and peace journalism, as outlined by prior work in conflict studies, in a corpus of news articles reporting on the Israel-Palestine war. For our analysis, we use computational approaches, using a combination of frame semantics and large language models to identify both communicative framing and its connection to linguistic framing. Our analysis reveals a higher focus on war based reporting rather than peace based. We also show substantial differences in reporting across the US, UK, and Middle Eastern news outlets in framing who the assailant and victims of the conflict are, surfacing biases within the media. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:20:54 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Kaur",
"Avneet",
""
],
[
"Arora",
"Arnav",
""
]
] |
2506.10422 | Mayank Patel | Mayank Patel and Minal Bhise | A Hybrid Heuristic Framework for Resource-Efficient Querying of Scientific Experiments Data | null | null | null | null | cs.DB cs.DC cs.ET cs.PF | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Scientific experiments and modern applications are generating large amounts of data every day. Most organizations utilize In-house servers or Cloud resources to manage application data and workload. The traditional database management system (DBMS) and HTAP systems spend significant time & resources to load the entire dataset into DBMS before starting query execution. On the other hand, in-situ engines may reparse required data multiple times, increasing resource utilization and data processing costs. Additionally, over or under-allocation of resources also increases application running costs. This paper proposes a lightweight Resource Availability &Workload aware Hybrid Framework (RAW-HF) to optimize querying raw data by utilizing existing finite resources efficiently. RAW-HF includes modules that help optimize the resources required to execute a given workload and maximize the utilization of existing resources. The impact of applying RAW-HF to real-world scientific dataset workloads like Sloan Digital Sky Survey (SDSS) and Linked Observation Data (LOD) presented over 90% and 85% reduction in workload execution time (WET) compared to widely used traditional DBMS PostgreSQL. The overall CPU, IO resource utilization, and WET have been reduced by 26%, 25%, and 26%, respectively, while improving memory utilization by 33%, compared to the state-of-the-art workload-aware partial loading technique (WA) proposed for hybrid systems. A comparison of MUAR technique used by RAW-HF with machine learning based resource allocation techniques like PCC is also presented. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:21:54 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Patel",
"Mayank",
""
],
[
"Bhise",
"Minal",
""
]
] |
2506.10423 | Tony Alex | Tony Alex, Wish Suharitdamrong, Sara Atito, Armin Mustafa, Philip J. B. Jackson, Imran Razzak, Muhammad Awais | PAL: Probing Audio Encoders via LLMs -- A Study of Information Transfer from Audio Encoders to LLMs | 21 pages, 11 figures | null | null | null | cs.SD cs.AI cs.CL eess.AS | http://creativecommons.org/licenses/by/4.0/ | The integration of audio perception capabilities into Large Language Models (LLMs) has enabled significant advances in Audio-LLMs. Although application-focused developments, particularly in curating training data for specific capabilities e.g., audio reasoning, have progressed rapidly, the underlying mechanisms that govern efficient transfer of rich semantic representations from audio encoders to LLMs remain under-explored. We conceptualize effective audio-LLM interaction as the LLM's ability to proficiently probe the audio encoder representations to satisfy textual queries. This paper presents a systematic investigation on how architectural design choices can affect that. Beginning with a standard Pengi/LLaVA-style audio-LLM architecture, we propose and evaluate several modifications guided by hypotheses derived from mechanistic interpretability studies and LLM operational principles. Our experiments demonstrate that: (1) delaying audio integration until the LLM's initial layers establish textual context that enhances its ability to probe the audio representations for relevant information; (2) the LLM can proficiently probe audio representations exclusively through LLM layer's attention submodule, without requiring propagation to its Feed-Forward Network (FFN) submodule; (3) an efficiently integrated ensemble of diverse audio encoders provides richer, complementary representations, thereby broadening the LLM's capacity to probe a wider spectrum of audio information. All hypotheses are evaluated using an identical three-stage training curriculum on a dataset of 5.6 million audio-text pairs, ensuring controlled comparisons. Our final architecture, which incorporates all proposed modifications, achieves relative improvements from 10\% to 60\% over the baseline, validating our approach to optimizing cross-modal information transfer in audio-LLMs. Project page: https://ta012.github.io/PAL/ | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:23:07 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Alex",
"Tony",
""
],
[
"Suharitdamrong",
"Wish",
""
],
[
"Atito",
"Sara",
""
],
[
"Mustafa",
"Armin",
""
],
[
"Jackson",
"Philip J. B.",
""
],
[
"Razzak",
"Imran",
""
],
[
"Awais",
"Muhammad",
""
]
] |
2506.10424 | Kaiyuan Zhang | Kaiyuan Zhang, Siyuan Cheng, Hanxi Guo, Yuetian Chen, Zian Su, Shengwei An, Yuntao Du, Charles Fleming, Ashish Kundu, Xiangyu Zhang, Ninghui Li | SOFT: Selective Data Obfuscation for Protecting LLM Fine-tuning against Membership Inference Attacks | Accepted by the 34th USENIX Security Symposium 2025. Code is available at https://github.com/KaiyuanZh/SOFT | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have achieved remarkable success and are widely adopted for diverse applications. However, fine-tuning these models often involves private or sensitive information, raising critical privacy concerns. In this work, we conduct the first comprehensive study evaluating the vulnerability of fine-tuned LLMs to membership inference attacks (MIAs). Our empirical analysis demonstrates that MIAs exploit the loss reduction during fine-tuning, making them highly effective in revealing membership information. These findings motivate the development of our defense. We propose SOFT (\textbf{S}elective data \textbf{O}bfuscation in LLM \textbf{F}ine-\textbf{T}uning), a novel defense technique that mitigates privacy leakage by leveraging influential data selection with an adjustable parameter to balance utility preservation and privacy protection. Our extensive experiments span six diverse domains and multiple LLM architectures and scales. Results show that SOFT effectively reduces privacy risks while maintaining competitive model performance, offering a practical and scalable solution to safeguard sensitive information in fine-tuned LLMs. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:23:56 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Kaiyuan",
""
],
[
"Cheng",
"Siyuan",
""
],
[
"Guo",
"Hanxi",
""
],
[
"Chen",
"Yuetian",
""
],
[
"Su",
"Zian",
""
],
[
"An",
"Shengwei",
""
],
[
"Du",
"Yuntao",
""
],
[
"Fleming",
"Charles",
""
],
[
"Kundu",
"Ashish",
""
],
[
"Zhang",
"Xiangyu",
""
],
[
"Li",
"Ninghui",
""
]
] |
2506.10425 | Guoyi Zhang | Guoyi Zhang, Guangsheng Xu, Siyang Chen, Han Wang and Xiaohu Zhang | It's Not the Target, It's the Background: Rethinking Infrared Small Target Detection via Deep Patch-Free Low-Rank Representations | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Infrared small target detection (IRSTD) remains a long-standing challenge in complex backgrounds due to low signal-to-clutter ratios (SCR), diverse target morphologies, and the absence of distinctive visual cues. While recent deep learning approaches aim to learn discriminative representations, the intrinsic variability and weak priors of small targets often lead to unstable performance. In this paper, we propose a novel end-to-end IRSTD framework, termed LRRNet, which leverages the low-rank property of infrared image backgrounds. Inspired by the physical compressibility of cluttered scenes, our approach adopts a compression--reconstruction--subtraction (CRS) paradigm to directly model structure-aware low-rank background representations in the image domain, without relying on patch-based processing or explicit matrix decomposition. To the best of our knowledge, this is the first work to directly learn low-rank background structures using deep neural networks in an end-to-end manner. Extensive experiments on multiple public datasets demonstrate that LRRNet outperforms 38 state-of-the-art methods in terms of detection accuracy, robustness, and computational efficiency. Remarkably, it achieves real-time performance with an average speed of 82.34 FPS. Evaluations on the challenging NoisySIRST dataset further confirm the model's resilience to sensor noise. The source code will be made publicly available upon acceptance. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:24:45 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Zhang",
"Guoyi",
""
],
[
"Xu",
"Guangsheng",
""
],
[
"Chen",
"Siyang",
""
],
[
"Wang",
"Han",
""
],
[
"Zhang",
"Xiaohu",
""
]
] |
2506.10426 | Haoxuan Chen | Xiao Yu, Haoxuan Chen, Feifei Niu, Xing Hu, Jacky Wai Keung, Xin Xia | Towards Understanding Bugs in Distributed Training and Inference Frameworks for Large Language Models | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rapid development of large language models (LLMs), distributed training and inference frameworks like DeepSpeed have become essential for scaling model training and inference across multiple GPUs or nodes. However, the increasing complexity of these frameworks brings non-trivial software bugs, which may degrade training performance, cause unexpected failures, and result in significant resource waste. Understanding framework bugs' characteristics is fundamental for quality assurance, allowing the design of more effective debugging and repair methods. Thus, our paper conducts the first large-scale empirical analysis of 308 fixed bugs across three popular distributed training/inference frameworks: DeepSpeed, Megatron-LM, and Colossal-AI. We examine bug symptoms, root causes, bug identification and fixing efforts, and common low-effort fixing strategies. Additionally, the distributed nature of these frameworks introduces unique bug root causes, such as allocation strategy error and distributed communication error. Diagnosing and fixing complex bugs remains challenging due to factors like the disconnect between symptoms and root causes, high bug reproduction costs, and low-level or cross-component interactions. Interestingly, we observe that 48% of bug fixes require minimal code changes (<=10 LOC) and follow simple strategies such as conditional logic optimization, parameter handling enhancement, or version compatibility handling, indicating potential for automation. Based on these insights, we offer several implications for improving the reliability of both distributed training and inference frameworks and their dependent LLM projects, while also identifying opportunities to leverage LLM-based tools for automated debugging and repair. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:24:59 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Yu",
"Xiao",
""
],
[
"Chen",
"Haoxuan",
""
],
[
"Niu",
"Feifei",
""
],
[
"Hu",
"Xing",
""
],
[
"Keung",
"Jacky Wai",
""
],
[
"Xia",
"Xin",
""
]
] |
2506.10427 | Nessrine Benaziz | Nessrine Benaziz, Thibaut Devolder and Jean-Paul Adam | Method of analysis of the spectra obtained by microfocused Brillouin light scattering | null | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | Brillouin Light Scattering is a powerful technique to measure the microwave excitations present in a magnetic system. In microfocused mode, the light is focused on the sample using a microscope objective. This accelerates substantially the measurement but results in mixing the response of all microwave excitations into a single spectrum, such that modeling is required to disentangle the contribution of each spin wave. By assuming that a spectrum collected in microfocused mode can be approximated by the sum of all back-scattering spectra that can be collected by the microscope objective, we develop a simple and direct way of interpreting spectra. The model is compared to experimental data collected on a 50 nm thick CoFeB magnetic film. The model allows the understanding of the influence of the optical properties of a sample, of the dispersion relation of the spin wave eigenexcitations and of their thickness profiles, as well as of their populations onto the magnitudes and the lineshapes of the characteristic features of a spectrum | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:26:54 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Benaziz",
"Nessrine",
""
],
[
"Devolder",
"Thibaut",
""
],
[
"Adam",
"Jean-Paul",
""
]
] |
2506.10428 | Sudeep Kundu | Sudeep Kundu and Shishu pal Singh | Penalty-Based Feedback Control and Finite Element Analysis for the Stabilization of Nonlinear Reaction-Diffusion Equations | null | null | null | null | math.NA cs.NA math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, first we employ the penalization technique to analyze the Dirichlet boundary feedback control problem pertaining to reaction-diffusion equation. We establish the stabilization result of the equivalent Robin problem in the \(H^{2}\)-norm with respect to the penalty parameter. Furthermore, we prove that the solution of the penalized control problem converges to the corresponding solution of the Dirichlet boundary feedback control problem as the penalty parameter \(\epsilon\) approaches zero. A \(C^{0}\)-conforming finite element method is applied to this problem for the spatial variable while keeping the time variable continuous. We discuss the stabilization of the semi-discrete scheme for the penalized control problem and present an error analysis of its solution. Finally, we validate our theoretical findings through numerical experiments. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:28:44 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Kundu",
"Sudeep",
""
],
[
"Singh",
"Shishu pal",
""
]
] |
2506.10429 | Kristoffer Simula | Kristoffer Simula, Maria-Andreea Filip, Ali Alavi | Transcorrelated Theory for Transition Metal Atoms | null | null | null | null | physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | We benchmark ionisation and excitation energies of transition-metal atoms Sc-Zn with a transcorrelated Hamiltonian combined with pseudopotentials. The similarity transformed Hamiltonian provides compact TC wave functions in affordable aug-cc-pVTZ and aug-cc-pVQZ Gaussian bases and eliminates the need for complete basis set extrapolations. The use of Douglas-Kroll-Hess theory is omitted because scalar relativistic effects are included in the pseudopotentials. Treating the full semicore (3s 3p) valence and freezing only 1s-2p shells, we reach chemical accuracy for all atoms and properties with coupled cluster and full configuration interaction quantum Monte Carlo. Consistent total energies across disparate orbital sets and correlation solvers highlights the robustness of the TC workflow. Our study pushes benchmark-quality quantum chemistry into the 3d block without large-scale basis sets and opens a practical route for transcorrelation to strongly correlated molecules and materials hosting heavier transition metals. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:32:31 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Simula",
"Kristoffer",
""
],
[
"Filip",
"Maria-Andreea",
""
],
[
"Alavi",
"Ali",
""
]
] |
2506.10430 | Jiahao Zhang | Shuo wang and Jihao Zhang | MF2Summ: Multimodal Fusion for Video Summarization with Temporal Alignment | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid proliferation of online video content necessitates effective video summarization techniques. Traditional methods, often relying on a single modality (typically visual), struggle to capture the full semantic richness of videos. This paper introduces MF2Summ, a novel video summarization model based on multimodal content understanding, integrating both visual and auditory information. MF2Summ employs a five-stage process: feature extraction, cross-modal attention interaction, feature fusion, segment prediction, and key shot selection. Visual features are extracted using a pre-trained GoogLeNet model, while auditory features are derived using SoundNet. The core of our fusion mechanism involves a cross-modal Transformer and an alignment-guided self-attention Transformer, designed to effectively model inter-modal dependencies and temporal correspondences. Segment importance, location, and center-ness are predicted, followed by key shot selection using Non-Maximum Suppression (NMS) and the Kernel Temporal Segmentation (KTS) algorithm. Experimental results on the SumMe and TVSum datasets demonstrate that MF2Summ achieves competitive performance, notably improving F1-scores by 1.9\% and 0.6\% respectively over the DSNet model, and performing favorably against other state-of-the-art methods. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:32:51 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"wang",
"Shuo",
""
],
[
"Zhang",
"Jihao",
""
]
] |
2506.10431 | Taichi Uemura | Taichi Uemura | Colimits in the $\infty$-category of $\infty$-topoi and \'etale morphisms | null | null | null | null | math.CT math.AT | http://creativecommons.org/licenses/by/4.0/ | We provide an alternative proof of Lurie's result that the wide subcategory of the $\infty$-category of $\infty$-topoi spanned by the \'etale morphisms is closed under small colimits. Our proof is based on a new characterization of \'etale morphisms of $\infty$-topoi in relation to univalent families and does not rely on a larger universe. During the proof, we also give an elementary construction of univalent completion. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:34:23 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Uemura",
"Taichi",
""
]
] |
2506.10432 | Vivek Kumar Anand | V. K. Anand, A. Bhattacharyya, D. T. Adroja, K. Panda, P. K. Biswas, A. D. Hillier, and B. Lake | Time reversal symmetry breaking and s-wave superconductivity in CaPd2Ge2: A $\mu$SR study | 8 pages, 5 figures | Phys. Rev. B 108, 224519 (2023) | 10.1103/PhysRevB.108.224519 | null | cond-mat.supr-con cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | ${\rm CaPd_2Ge_2}$ which crystallizes in ${\rm ThCr_2Si_2}$-type body-centered tetragonal structure exhibits superconductivity below the critical temperature $T_{\rm c} = 1.69$~K\@. We have investigated the superconducting gap structure and time reversal symmetry of the ground state in ${\rm CaPd_2Ge_2}$ by means of muon spin relaxation and rotation ($\mu$SR) measurements. Our analysis of $\mu$SR data collected in transverse magnetic field reveals BCS superconductivity with a single-band $s$-wave singlet pairing and an isotropic energy gap having the value $2\Delta(0)/k_{\rm B}T_{\rm c} = 3.50(1)$. Further, an increased relaxation rate in zero field $\mu$SR asymmetry spectra below $T_{\rm c} $ provides evidence for the presence of a spontaneous magnetic field in the superconducting state revealing that the time-reversal symmetry is broken in ${\rm CaPd_2Ge_2}$. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:35:25 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Anand",
"V. K.",
""
],
[
"Bhattacharyya",
"A.",
""
],
[
"Adroja",
"D. T.",
""
],
[
"Panda",
"K.",
""
],
[
"Biswas",
"P. K.",
""
],
[
"Hillier",
"A. D.",
""
],
[
"Lake",
"B.",
""
]
] |
2506.10433 | Florian Handke | Florian Handke, F\'elix Koulischer, Gabriel Raya, Luca Ambrogioni | Measuring Semantic Information Production in Generative Diffusion Models | 4 pages, 3 figures, an appendix with derivations and implementation details, accepted at ICLR DeLTa 2025 | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | It is well known that semantic and structural features of the generated images emerge at different times during the reverse dynamics of diffusion, a phenomenon that has been connected to physical phase transitions in magnets and other materials. In this paper, we introduce a general information-theoretic approach to measure when these class-semantic "decisions" are made during the generative process. By using an online formula for the optimal Bayesian classifier, we estimate the conditional entropy of the class label given the noisy state. We then determine the time intervals corresponding to the highest information transfer between noisy states and class labels using the time derivative of the conditional entropy. We demonstrate our method on one-dimensional Gaussian mixture models and on DDPM models trained on the CIFAR10 dataset. As expected, we find that the semantic information transfer is highest in the intermediate stages of diffusion while vanishing during the final stages. However, we found sizable differences between the entropy rate profiles of different classes, suggesting that different "semantic decisions" are located at different intermediate times. | [
{
"version": "v1",
"created": "Thu, 12 Jun 2025 07:35:29 GMT"
}
] | 2025-06-13T00:00:00 | [
[
"Handke",
"Florian",
""
],
[
"Koulischer",
"Félix",
""
],
[
"Raya",
"Gabriel",
""
],
[
"Ambrogioni",
"Luca",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.