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
list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2508.09998
|
Lucie-Aim\'ee Kaffee
|
Lucie-Aim\'ee Kaffee and Giada Pistilli and Yacine Jernite
|
INTIMA: A Benchmark for Human-AI Companionship Behavior
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
AI companionship, where users develop emotional bonds with AI systems, has emerged as a significant pattern with positive but also concerning implications. We introduce Interactions and Machine Attachment Benchmark (INTIMA), a benchmark for evaluating companionship behaviors in language models. Drawing from psychological theories and user data, we develop a taxonomy of 31 behaviors across four categories and 368 targeted prompts. Responses to these prompts are evaluated as companionship-reinforcing, boundary-maintaining, or neutral. Applying INTIMA to Gemma-3, Phi-4, o3-mini, and Claude-4 reveals that companionship-reinforcing behaviors remain much more common across all models, though we observe marked differences between models. Different commercial providers prioritize different categories within the more sensitive parts of the benchmark, which is concerning since both appropriate boundary-setting and emotional support matter for user well-being. These findings highlight the need for more consistent approaches to handling emotionally charged interactions.
|
[
{
"version": "v1",
"created": "Mon, 04 Aug 2025 08:25:38 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Kaffee",
"Lucie-Aimée",
""
],
[
"Pistilli",
"Giada",
""
],
[
"Jernite",
"Yacine",
""
]
] |
2508.09999
|
Yuzhuo Xiao
|
Yuzhuo Xiao, Zeyu Han, Yuhan Wang, Huaizu Jiang
|
XFacta: Contemporary, Real-World Dataset and Evaluation for Multimodal Misinformation Detection with Multimodal LLMs
|
For associated code and dataset, see https://github.com/neu-vi/XFacta
| null | null | null |
cs.CL cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The rapid spread of multimodal misinformation on social media calls for more effective and robust detection methods. Recent advances leveraging multimodal large language models (MLLMs) have shown the potential in addressing this challenge. However, it remains unclear exactly where the bottleneck of existing approaches lies (evidence retrieval v.s. reasoning), hindering the further advances in this field. On the dataset side, existing benchmarks either contain outdated events, leading to evaluation bias due to discrepancies with contemporary social media scenarios as MLLMs can simply memorize these events, or artificially synthetic, failing to reflect real-world misinformation patterns. Additionally, it lacks comprehensive analyses of MLLM-based model design strategies. To address these issues, we introduce XFacta, a contemporary, real-world dataset that is better suited for evaluating MLLM-based detectors. We systematically evaluate various MLLM-based misinformation detection strategies, assessing models across different architectures and scales, as well as benchmarking against existing detection methods. Building on these analyses, we further enable a semi-automatic detection-in-the-loop framework that continuously updates XFacta with new content to maintain its contemporary relevance. Our analysis provides valuable insights and practices for advancing the field of multimodal misinformation detection. The code and data have been released.
|
[
{
"version": "v1",
"created": "Mon, 04 Aug 2025 14:14:52 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Xiao",
"Yuzhuo",
""
],
[
"Han",
"Zeyu",
""
],
[
"Wang",
"Yuhan",
""
],
[
"Jiang",
"Huaizu",
""
]
] |
2508.10000
|
Chenhao Xue
|
Chenhao Xue, Yuanzhe Jin, Adrian Carrasco-Revilla, Joyraj Chakraborty, Min Chen
|
AutoGeTS: Knowledge-based Automated Generation of Text Synthetics for Improving Text Classification
| null | null | null | null |
cs.CL cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
When developing text classification models for real world applications, one major challenge is the difficulty to collect sufficient data for all text classes. In this work, we address this challenge by utilizing large language models (LLMs) to generate synthetic data and using such data to improve the performance of the models without waiting for more real data to be collected and labelled. As an LLM generates different synthetic data in response to different input examples, we formulate an automated workflow, which searches for input examples that lead to more ``effective'' synthetic data for improving the model concerned. We study three search strategies with an extensive set of experiments, and use experiment results to inform an ensemble algorithm that selects a search strategy according to the characteristics of a class. Our further experiments demonstrate that this ensemble approach is more effective than each individual strategy in our automated workflow for improving classification models using LLMs.
|
[
{
"version": "v1",
"created": "Mon, 04 Aug 2025 16:53:20 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Xue",
"Chenhao",
""
],
[
"Jin",
"Yuanzhe",
""
],
[
"Carrasco-Revilla",
"Adrian",
""
],
[
"Chakraborty",
"Joyraj",
""
],
[
"Chen",
"Min",
""
]
] |
2508.10001
|
Gauri Chopra
|
Rakesh Thakur, Sneha Sharma, Gauri Chopra
|
HiFACTMix: A Code-Mixed Benchmark and Graph-Aware Model for EvidenceBased Political Claim Verification in Hinglish
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Fact-checking in code-mixed, low-resource languages such as Hinglish remains an underexplored challenge in natural language processing. Existing fact-verification systems largely focus on high-resource, monolingual settings and fail to generalize to real-world political discourse in linguistically diverse regions like India. Given the widespread use of Hinglish by public figures, particularly political figures, and the growing influence of social media on public opinion, there's a critical need for robust, multilingual and context-aware fact-checking tools. To address this gap a novel benchmark HiFACT dataset is introduced with 1,500 realworld factual claims made by 28 Indian state Chief Ministers in Hinglish, under a highly code-mixed low-resource setting. Each claim is annotated with textual evidence and veracity labels. To evaluate this benchmark, a novel graphaware, retrieval-augmented fact-checking model is proposed that combines multilingual contextual encoding, claim-evidence semantic alignment, evidence graph construction, graph neural reasoning, and natural language explanation generation. Experimental results show that HiFACTMix outperformed accuracy in comparison to state of art multilingual baselines models and provides faithful justifications for its verdicts. This work opens a new direction for multilingual, code-mixed, and politically grounded fact verification research.
|
[
{
"version": "v1",
"created": "Mon, 04 Aug 2025 17:14:03 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Thakur",
"Rakesh",
""
],
[
"Sharma",
"Sneha",
""
],
[
"Chopra",
"Gauri",
""
]
] |
2508.10002
|
Geanderson Carvalho
|
G.A. Carvalho and J.D.V. Arba\~nil and J.G. Coelho
|
Effects of light-mass fermionic dark matter on the equilibrium and stability of white dwarfs
| null | null | null | null |
astro-ph.HE astro-ph.SR gr-qc hep-ph hep-th
|
http://creativecommons.org/licenses/by/4.0/
|
White dwarfs (WDs) can be used as laboratories to test strong gravity and high-density regimes, once their equation of state is not so uncertain as the one of neutron stars. This makes them also a useful tool to constrain dark-matter models. In this work, we study dark matter white dwarfs (DMWD) composed of white dwarf matter admixed with fermionic dark matter in a two-fluid general relativistic framework. Dark matter particles are considered to have masses between $0.1-10$ GeV. The equilibrium configurations and stability are derived, showing that the DMWD can be more compact, with masses around 1.3 $M\odot$ and radii around 500 km. The increasing compactness leads to changes in the fundamental modes of radial oscillations ($\sim20\%$ for 0.1 GeV DM), which produce detectable shifts in GW frequencies. The interplay between dark matter and normal matter thus provides a compelling avenue for interpreting deviations in observed WD properties and for placing constraints on DM characteristics through astrophysical observations.
|
[
{
"version": "v1",
"created": "Mon, 04 Aug 2025 18:59:27 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Carvalho",
"G. A.",
""
],
[
"Arbañil",
"J. D. V.",
""
],
[
"Coelho",
"J. G.",
""
]
] |
2508.10003
|
Austin Kozlowski
|
Austin C. Kozlowski, Callin Dai, Andrei Boutyline
|
Semantic Structure in Large Language Model Embeddings
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Psychological research consistently finds that human ratings of words across diverse semantic scales can be reduced to a low-dimensional form with relatively little information loss. We find that the semantic associations encoded in the embedding matrices of large language models (LLMs) exhibit a similar structure. We show that the projections of words on semantic directions defined by antonym pairs (e.g. kind - cruel) correlate highly with human ratings, and further find that these projections effectively reduce to a 3-dimensional subspace within LLM embeddings, closely resembling the patterns derived from human survey responses. Moreover, we find that shifting tokens along one semantic direction causes off-target effects on geometrically aligned features proportional to their cosine similarity. These findings suggest that semantic features are entangled within LLMs similarly to how they are interconnected in human language, and a great deal of semantic information, despite its apparent complexity, is surprisingly low-dimensional. Furthermore, accounting for this semantic structure may prove essential for avoiding unintended consequences when steering features.
|
[
{
"version": "v1",
"created": "Mon, 04 Aug 2025 20:21:50 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Kozlowski",
"Austin C.",
""
],
[
"Dai",
"Callin",
""
],
[
"Boutyline",
"Andrei",
""
]
] |
2508.10004
|
Andres Carvallo
|
Andr\'es Carvallo, Denis Parra, Peter Brusilovsky, Hernan Valdivieso, Gabriel Rada, Ivania Donoso and Vladimir Araujo
|
User Perception of Attention Visualizations: Effects on Interpretability Across Evidence-Based Medical Documents
| null | null | null | null |
cs.CL cs.AI cs.HC cs.IR cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
The attention mechanism is a core component of the Transformer architecture. Beyond improving performance, attention has been proposed as a mechanism for explainability via attention weights, which are associated with input features (e.g., tokens in a document). In this context, larger attention weights may imply more relevant features for the model's prediction. In evidence-based medicine, such explanations could support physicians' understanding and interaction with AI systems used to categorize biomedical literature. However, there is still no consensus on whether attention weights provide helpful explanations. Moreover, little research has explored how visualizing attention affects its usefulness as an explanation aid. To bridge this gap, we conducted a user study to evaluate whether attention-based explanations support users in biomedical document classification and whether there is a preferred way to visualize them. The study involved medical experts from various disciplines who classified articles based on study design (e.g., systematic reviews, broad synthesis, randomized and non-randomized trials). Our findings show that the Transformer model (XLNet) classified documents accurately; however, the attention weights were not perceived as particularly helpful for explaining the predictions. However, this perception varied significantly depending on how attention was visualized. Contrary to Munzner's principle of visual effectiveness, which favors precise encodings like bar length, users preferred more intuitive formats, such as text brightness or background color. While our results do not confirm the overall utility of attention weights for explanation, they suggest that their perceived helpfulness is influenced by how they are visually presented.
|
[
{
"version": "v1",
"created": "Tue, 05 Aug 2025 13:24:52 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Carvallo",
"Andrés",
""
],
[
"Parra",
"Denis",
""
],
[
"Brusilovsky",
"Peter",
""
],
[
"Valdivieso",
"Hernan",
""
],
[
"Rada",
"Gabriel",
""
],
[
"Donoso",
"Ivania",
""
],
[
"Araujo",
"Vladimir",
""
]
] |
2508.10005
|
Chengliang Zhou
|
Chengliang Zhou, Mei Wang, Ting Zhang, Qiannan Zhu, Jian Li, Hua Huang
|
From Answers to Questions: EQGBench for Evaluating LLMs' Educational Question Generation
| null | null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large Language Models (LLMs) have demonstrated remarkable capabilities in mathematical problem-solving. However, the transition from providing answers to generating high-quality educational questions presents significant challenges that remain underexplored. To advance Educational Question Generation (EQG) and facilitate LLMs in generating pedagogically valuable and educationally effective questions, we introduce EQGBench, a comprehensive benchmark specifically designed for evaluating LLMs' performance in Chinese EQG. EQGBench establishes a five-dimensional evaluation framework supported by a dataset of 900 evaluation samples spanning three fundamental middle school disciplines: mathematics, physics, and chemistry. The dataset incorporates user queries with varying knowledge points, difficulty gradients, and question type specifications to simulate realistic educational scenarios. Through systematic evaluation of 46 mainstream large models, we reveal significant room for development in generating questions that reflect educational value and foster students' comprehensive abilities.
|
[
{
"version": "v1",
"created": "Tue, 05 Aug 2025 14:16:42 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zhou",
"Chengliang",
""
],
[
"Wang",
"Mei",
""
],
[
"Zhang",
"Ting",
""
],
[
"Zhu",
"Qiannan",
""
],
[
"Li",
"Jian",
""
],
[
"Huang",
"Hua",
""
]
] |
2508.10006
|
Adam Brownstein
|
Adam Brownstein
|
On something and nothing: The interface of two types of nothing
| null | null | null | null |
physics.hist-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We suggest that the question of why is there something rather than nothing can be answered by the existence of two types of nothing. We propose that matter occurs at the boundaries of intersection of both nothings. This accords with the common view that there are three worlds; the platonic world of concepts, the material world (i.e. of matter), and the non-material world (i.e. of consciousness). Both the material and non-material worlds have their own type of nothing, thus leading to the proposal. The interpretation provides an alternative type of duality distinct from property or substance dualism, and may unify the two. The interpretation also has implications for the understanding of physical causation, and Mach's principle as the boundary of intersection provides a return to the aether concept.
|
[
{
"version": "v1",
"created": "Tue, 05 Aug 2025 17:00:37 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Brownstein",
"Adam",
""
]
] |
2508.10007
|
Yizhou Lyu
|
Y. Lyu, D. Combs, D. Neumann, Y. C. Leong
|
Automated scoring of the Ambiguous Intentions Hostility Questionnaire using fine-tuned large language models
|
We have no known conflict of interest
| null | null | null |
cs.CL stat.ME
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Hostile attribution bias is the tendency to interpret social interactions as intentionally hostile. The Ambiguous Intentions Hostility Questionnaire (AIHQ) is commonly used to measure hostile attribution bias, and includes open-ended questions where participants describe the perceived intentions behind a negative social situation and how they would respond. While these questions provide insights into the contents of hostile attributions, they require time-intensive scoring by human raters. In this study, we assessed whether large language models can automate the scoring of AIHQ open-ended responses. We used a previously collected dataset in which individuals with traumatic brain injury (TBI) and healthy controls (HC) completed the AIHQ and had their open-ended responses rated by trained human raters. We used half of these responses to fine-tune the two models on human-generated ratings, and tested the fine-tuned models on the remaining half of AIHQ responses. Results showed that model-generated ratings aligned with human ratings for both attributions of hostility and aggression responses, with fine-tuned models showing higher alignment. This alignment was consistent across ambiguous, intentional, and accidental scenario types, and replicated previous findings on group differences in attributions of hostility and aggression responses between TBI and HC groups. The fine-tuned models also generalized well to an independent nonclinical dataset. To support broader adoption, we provide an accessible scoring interface that includes both local and cloud-based options. Together, our findings suggest that large language models can streamline AIHQ scoring in both research and clinical contexts, revealing their potential to facilitate psychological assessments across different populations.
|
[
{
"version": "v1",
"created": "Tue, 05 Aug 2025 21:58:11 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Lyu",
"Y.",
""
],
[
"Combs",
"D.",
""
],
[
"Neumann",
"D.",
""
],
[
"Leong",
"Y. C.",
""
]
] |
2508.10008
|
Antonio Leandro Martins Candido Mr
|
Antonio Leandro Martins Candido and Jose Everardo Bessa Maia
|
Multidimensional classification of posts for online course discussion forum curation
|
8 pages, 1 figure
| null | null | null |
cs.CL cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
The automatic curation of discussion forums in online courses requires constant updates, making frequent retraining of Large Language Models (LLMs) a resource-intensive process. To circumvent the need for costly fine-tuning, this paper proposes and evaluates the use of Bayesian fusion. The approach combines the multidimensional classification scores of a pre-trained generic LLM with those of a classifier trained on local data. The performance comparison demonstrated that the proposed fusion improves the results compared to each classifier individually, and is competitive with the LLM fine-tuning approach
|
[
{
"version": "v1",
"created": "Tue, 05 Aug 2025 22:53:01 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Candido",
"Antonio Leandro Martins",
""
],
[
"Maia",
"Jose Everardo Bessa",
""
]
] |
2508.10009
|
Eunsoo Hong
|
Hojun Jin, Eunsoo Hong, Ziwon Hyung, Sungjun Lim, Seungjin Lee, Keunseok Cho
|
Beyond Hard Sharing: Efficient Multi-Task Speech-to-Text Modeling with Supervised Mixture of Experts
|
Accepted to Interspeech 2025
| null | null | null |
cs.CL cs.AI cs.SD eess.AS
|
http://creativecommons.org/licenses/by/4.0/
|
Hard-parameter sharing is a common strategy to train a single model jointly across diverse tasks. However, this often leads to task interference, impeding overall model performance. To address the issue, we propose a simple yet effective Supervised Mixture of Experts (S-MoE). Unlike traditional Mixture of Experts models, S-MoE eliminates the need for training gating functions by utilizing special guiding tokens to route each task to its designated expert. By assigning each task to a separate feedforward network, S-MoE overcomes the limitations of hard-parameter sharing. We further apply S-MoE to a speech-to-text model, enabling the model to process mixed-bandwidth input while jointly performing automatic speech recognition (ASR) and speech translation (ST). Experimental results demonstrate the effectiveness of the proposed S-MoE, achieving a 6.35% relative improvement in Word Error Rate (WER) when applied to both the encoder and decoder.
|
[
{
"version": "v1",
"created": "Tue, 05 Aug 2025 23:56:11 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Jin",
"Hojun",
""
],
[
"Hong",
"Eunsoo",
""
],
[
"Hyung",
"Ziwon",
""
],
[
"Lim",
"Sungjun",
""
],
[
"Lee",
"Seungjin",
""
],
[
"Cho",
"Keunseok",
""
]
] |
2508.10010
|
Ayana Hussain
|
Ayana Hussain, Patrick Zhao and Nicholas Vincent
|
An Audit and Analysis of LLM-Assisted Health Misinformation Jailbreaks Against LLMs
| null | null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Large Language Models (LLMs) are a double-edged sword capable of generating harmful misinformation -- inadvertently, or when prompted by "jailbreak" attacks that attempt to produce malicious outputs. LLMs could, with additional research, be used to detect and prevent the spread of misinformation. In this paper, we investigate the efficacy and characteristics of LLM-produced jailbreak attacks that cause other models to produce harmful medical misinformation. We also study how misinformation generated by jailbroken LLMs compares to typical misinformation found on social media, and how effectively it can be detected using standard machine learning approaches. Specifically, we closely examine 109 distinct attacks against three target LLMs and compare the attack prompts to in-the-wild health-related LLM queries. We also examine the resulting jailbreak responses, comparing the generated misinformation to health-related misinformation on Reddit. Our findings add more evidence that LLMs can be effectively used to detect misinformation from both other LLMs and from people, and support a body of work suggesting that with careful design, LLMs can contribute to a healthier overall information ecosystem.
|
[
{
"version": "v1",
"created": "Wed, 06 Aug 2025 02:14:28 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Hussain",
"Ayana",
""
],
[
"Zhao",
"Patrick",
""
],
[
"Vincent",
"Nicholas",
""
]
] |
2508.10011
|
Licht Miyamoto
|
Yuta Nagamori, Mikoto Kosai, Yuji Kawai, Haruka Marumo, Misaki Shibuya, Tatsuya Negishi, Masaki Imanishi, Yasumasa Ikeda, Koichiro Tsuchiya, Asuka Sawai, Licht Miyamoto
|
Evaluation of GPT-based large language generative AI models as study aids for the national licensure examination for registered dietitians in Japan
| null | null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Generative artificial intelligence (AI) based on large language models (LLMs), such as ChatGPT, has demonstrated remarkable progress across various professional fields, including medicine and education. However, their performance in nutritional education, especially in Japanese national licensure examination for registered dietitians, remains underexplored. This study aimed to evaluate the potential of current LLM-based generative AI models as study aids for nutrition students. Questions from the Japanese national examination for registered dietitians were used as prompts for ChatGPT and three Bing models (Precise, Creative, Balanced), based on GPT-3.5 and GPT-4. Each question was entered into independent sessions, and model responses were analyzed for accuracy, consistency, and response time. Additional prompt engineering, including role assignment, was tested to assess potential performance improvements. Bing-Precise (66.2%) and Bing-Creative (61.4%) surpassed the passing threshold (60%), while Bing-Balanced (43.3%) and ChatGPT (42.8%) did not. Bing-Precise and Bing-Creative generally outperformed others across subject fields except Nutrition Education, where all models underperformed. None of the models consistently provided the same correct responses across repeated attempts, highlighting limitations in answer stability. ChatGPT showed greater consistency in response patterns but lower accuracy. Prompt engineering had minimal effect, except for modest improvement when correct answers and explanations were explicitly provided. While some generative AI models marginally exceeded the passing threshold, overall accuracy and answer consistency remained suboptimal. Moreover, all the models demonstrated notable limitations in answer consistency and robustness. Further advancements are needed to ensure reliable and stable AI-based study aids for dietitian licensure preparation.
|
[
{
"version": "v1",
"created": "Tue, 05 Aug 2025 03:33:11 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Nagamori",
"Yuta",
""
],
[
"Kosai",
"Mikoto",
""
],
[
"Kawai",
"Yuji",
""
],
[
"Marumo",
"Haruka",
""
],
[
"Shibuya",
"Misaki",
""
],
[
"Negishi",
"Tatsuya",
""
],
[
"Imanishi",
"Masaki",
""
],
[
"Ikeda",
"Yasumasa",
""
],
[
"Tsuchiya",
"Koichiro",
""
],
[
"Sawai",
"Asuka",
""
],
[
"Miyamoto",
"Licht",
""
]
] |
2508.10012
|
Dehao Tao
|
Dehao Tao, Guangjie Liu, Weizheng, Yongfeng Huang, Minghu jiang
|
Guided Navigation in Knowledge-Dense Environments: Structured Semantic Exploration with Guidance Graphs
| null | null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
While Large Language Models (LLMs) exhibit strong linguistic capabilities, their reliance on static knowledge and opaque reasoning processes limits their performance in knowledge intensive tasks. Knowledge graphs (KGs) offer a promising solution, but current exploration methods face a fundamental trade off: question guided approaches incur redundant exploration due to granularity mismatches, while clue guided methods fail to effectively leverage contextual information for complex scenarios. To address these limitations, we propose Guidance Graph guided Knowledge Exploration (GG Explore), a novel framework that introduces an intermediate Guidance Graph to bridge unstructured queries and structured knowledge retrieval. The Guidance Graph defines the retrieval space by abstracting the target knowledge' s structure while preserving broader semantic context, enabling precise and efficient exploration. Building upon the Guidance Graph, we develop: (1) Structural Alignment that filters incompatible candidates without LLM overhead, and (2) Context Aware Pruning that enforces semantic consistency with graph constraints. Extensive experiments show our method achieves superior efficiency and outperforms SOTA, especially on complex tasks, while maintaining strong performance with smaller LLMs, demonstrating practical value.
|
[
{
"version": "v1",
"created": "Wed, 06 Aug 2025 08:47:57 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Tao",
"Dehao",
""
],
[
"Liu",
"Guangjie",
""
],
[
"Weizheng",
"",
""
],
[
"Huang",
"Yongfeng",
""
],
[
"jiang",
"Minghu",
""
]
] |
2508.10013
|
Linqing Chen
|
Linqing Chen, Hanmeng Zhong, Wentao Wu, Weilei Wang
|
Semantic Bridge: Universal Multi-Hop Question Generation via AMR-Driven Graph Synthesis
| null | null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Large language model (LLM) training faces a critical bottleneck: the scarcity of high-quality, reasoning-intensive question-answer pairs, especially from sparse, domain-specific sources like PubMed papers or legal documents. Existing methods rely on surface patterns, fundamentally failing to generate controllable, complex multi-hop reasoning questions that test genuine understanding-essential for advancing LLM training paradigms. We present \textbf{Semantic Bridge}, the first universal framework for controllably generating sophisticated multi-hop reasoning questions from arbitrary sources. Our breakthrough innovation is \textit{semantic graph weaving}-three complementary bridging mechanisms (entity bridging for role-varying shared entities, predicate chain bridging for temporal/causal/logical sequences, and causal bridging for explicit reasoning chains)-that systematically construct complex pathways across documents, with fine-grained control over complexity and types via AMR-driven analysis. Our multi-modal AMR pipeline achieves up to 9.5% better round-trip quality, enabling production-ready controllable QA generation. Extensive evaluation demonstrates performance across both general-purpose datasets (Wikipedia) and specialized domains (biomedicine) It yields consistent 18.3%-25.4% gains over baselines across four languages (English, Chinese, French, German). Question pairs generated from 200 sources outperform 600 native human annotation examples with 67% fewer materials. Human evaluation shows 23.4% higher complexity, 18.7% better answerability, and 31.2% improved pattern coverage. Semantic Bridge establishes a new paradigm for LLM training data synthesis, enabling controllable generation of targeted reasoning questions from sparse sources. We will release our core code and semantic bridge model.
|
[
{
"version": "v1",
"created": "Wed, 06 Aug 2025 10:59:42 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Chen",
"Linqing",
""
],
[
"Zhong",
"Hanmeng",
""
],
[
"Wu",
"Wentao",
""
],
[
"Wang",
"Weilei",
""
]
] |
2508.10014
|
Lingfeng Zhou
|
Lingfeng Zhou, Jialing Zhang, Jin Gao, Mohan Jiang, Dequan Wang
|
PersonaEval: Are LLM Evaluators Human Enough to Judge Role-Play?
|
Accepted by COLM 2025
| null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Current role-play studies often rely on unvalidated LLM-as-a-judge paradigms, which may fail to reflect how humans perceive role fidelity. A key prerequisite for human-aligned evaluation is role identification, the ability to recognize who is speaking based on dialogue context. We argue that any meaningful judgment of role-playing quality (how well a character is played) fundamentally depends on first correctly attributing words and actions to the correct persona (who is speaking). We present PersonaEval, the first benchmark designed to test whether LLM evaluators can reliably identify human roles. PersonaEval uses human-authored dialogues from novels, scripts, and video transcripts, challenging models to determine the correct persona according to the conversation context. Our experiments, including a human study, show that even the best-performing LLMs reach only around 69% accuracy, well below the level needed for reliable evaluation. In contrast, human participants perform near ceiling with 90.8% accuracy, highlighting that current LLM evaluators are still not human enough to effectively judge role-play scenarios. To better understand this gap, we examine training-time adaptation and test-time compute, suggesting that reliable evaluation requires more than task-specific tuning, but depends on strong, human-like reasoning abilities in LLM evaluators. We release our benchmark at https://github.com/maple-zhou/PersonaEval.
|
[
{
"version": "v1",
"created": "Wed, 06 Aug 2025 13:06:15 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zhou",
"Lingfeng",
""
],
[
"Zhang",
"Jialing",
""
],
[
"Gao",
"Jin",
""
],
[
"Jiang",
"Mohan",
""
],
[
"Wang",
"Dequan",
""
]
] |
2508.10015
|
Enzhi Wang
|
Enzhi Wang, Qicheng Li, Shiwan Zhao, Aobo Kong, Jiaming Zhou, Xi Yang, Yequan Wang, Yonghua Lin, Yong Qin
|
RealTalk-CN: A Realistic Chinese Speech-Text Dialogue Benchmark With Cross-Modal Interaction Analysis
|
9 pages
| null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In recent years, large language models (LLMs) have achieved remarkable advancements in multimodal processing, including end-to-end speech-based language models that enable natural interactions and perform specific tasks in task-oriented dialogue (TOD) systems. However, existing TOD datasets are predominantly text-based, lacking real speech signals that are essential for evaluating the robustness of speech-based LLMs. Moreover, existing speech TOD datasets are primarily English and lack critical aspects such as speech disfluencies and speaker variations. To address these gaps, we introduce RealTalk-CN, the first Chinese multi-turn, multi-domain speech-text dual-modal TOD dataset, comprising 5.4k dialogues (60K utterances, 150 hours) with paired speech-text annotations. RealTalk-CN captures diverse dialogue scenarios with annotated spontaneous speech disfluencies, ensuring comprehensive coverage of real-world complexities in speech dialogue. In addition, we propose a novel cross-modal chat task that authentically simulates real-world user interactions, allowing dynamic switching between speech and text modalities. Our evaluation covers robustness to speech disfluencies, sensitivity to speaker characteristics, and cross-domain performance. Extensive experiments validate the effectiveness of RealTalk-CN, establishing a strong foundation for Chinese speech-based LLMs research.
|
[
{
"version": "v1",
"created": "Wed, 06 Aug 2025 13:12:57 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Wang",
"Enzhi",
""
],
[
"Li",
"Qicheng",
""
],
[
"Zhao",
"Shiwan",
""
],
[
"Kong",
"Aobo",
""
],
[
"Zhou",
"Jiaming",
""
],
[
"Yang",
"Xi",
""
],
[
"Wang",
"Yequan",
""
],
[
"Lin",
"Yonghua",
""
],
[
"Qin",
"Yong",
""
]
] |
2508.10016
|
Tianyu Xie
|
Tianyu Xie, Yuhang Wu, Yongdong Luo, Jiayi Ji, and Xiawu Zheng
|
Training-Free Multimodal Large Language Model Orchestration
| null | null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Different Multimodal Large Language Models (MLLMs) cannot be integrated into a unified multimodal input-output system directly. In previous work, training has been considered as an inevitable component due to challenges in modal alignment, Text-to-Speech efficiency and other integration issues. In this paper, we introduce Multimodal Large Language Model Orchestration, an effective approach for creating interactive multimodal AI systems without additional training. MLLM Orchestration leverages the inherent reasoning capabilities of large language models to coordinate specialized models through explicit workflows, enabling natural multimodal interactions while maintaining modularity, improving interpretability, and significantly enhancing computational efficiency. Our orchestration framework is built upon three key innovations: (1) a central controller LLM that analyzes user inputs and dynamically routes tasks to appropriate specialized models through carefully designed agents; (2) a parallel Text-to-Speech architecture that enables true full-duplex interaction with seamless interruption handling and natural conversational flow; and (3) a cross-modal memory integration system that maintains coherent context across modalities through intelligent information synthesis and retrieval, selectively avoiding unnecessary modality calls in certain scenarios to improve response speed. Extensive evaluations demonstrate that MLLM Orchestration achieves comprehensive multimodal capabilities without additional training, performance improvements of up to 7.8% over traditional jointly-trained approaches on standard benchmarks, reduced latency by 10.3%, and significantly enhanced interpretability through explicit orchestration processes.
|
[
{
"version": "v1",
"created": "Wed, 06 Aug 2025 16:17:29 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Xie",
"Tianyu",
""
],
[
"Wu",
"Yuhang",
""
],
[
"Luo",
"Yongdong",
""
],
[
"Ji",
"Jiayi",
""
],
[
"Zheng",
"Xiawu",
""
]
] |
2508.10017
|
Rodrigo Tertulino
|
Rodrigo Tertulino
|
A Robust Pipeline for Differentially Private Federated Learning on Imbalanced Clinical Data using SMOTETomek and FedProx
|
This is being prepared to be submitted to the Journal of the Brazilian Computer Society (JBCS), which is still under construction
| null | null | null |
cs.CR cs.AI cs.LG cs.SE
|
http://creativecommons.org/licenses/by/4.0/
|
Federated Learning (FL) presents a groundbreaking approach for collaborative health research, allowing model training on decentralized data while safeguarding patient privacy. FL offers formal security guarantees when combined with Differential Privacy (DP). The integration of these technologies, however, introduces a significant trade-off between privacy and clinical utility, a challenge further complicated by the severe class imbalance often present in medical datasets. The research presented herein addresses these interconnected issues through a systematic, multi-stage analysis. An FL framework was implemented for cardiovascular risk prediction, where initial experiments showed that standard methods struggled with imbalanced data, resulting in a recall of zero. To overcome such a limitation, we first integrated the hybrid Synthetic Minority Over-sampling Technique with Tomek Links (SMOTETomek) at the client level, successfully developing a clinically useful model. Subsequently, the framework was optimized for non-IID data using a tuned FedProx algorithm. Our final results reveal a clear, non-linear trade-off between the privacy budget (epsilon) and model recall, with the optimized FedProx consistently out-performing standard FedAvg. An optimal operational region was identified on the privacy-utility frontier, where strong privacy guarantees (with epsilon 9.0) can be achieved while maintaining high clinical utility (recall greater than 77%). Ultimately, our study provides a practical methodological blueprint for creating effective, secure, and accurate diagnostic tools that can be applied to real-world, heterogeneous healthcare data.
|
[
{
"version": "v1",
"created": "Wed, 06 Aug 2025 20:47:50 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Tertulino",
"Rodrigo",
""
]
] |
2508.10018
|
Sridhar Mahadevan
|
Sridhar Mahadevan
|
A Rose by Any Other Name Would Smell as Sweet: Categorical Homotopy Theory for Large Language Models
|
26 pages. arXiv admin note: text overlap with arXiv:2402.18732
| null | null | null |
cs.CL cs.AI math.AT
|
http://creativecommons.org/licenses/by/4.0/
|
Natural language is replete with superficially different statements, such as ``Charles Darwin wrote" and ``Charles Darwin is the author of", which carry the same meaning. Large language models (LLMs) should generate the same next-token probabilities in such cases, but usually do not. Empirical workarounds have been explored, such as using k-NN estimates of sentence similarity to produce smoothed estimates. In this paper, we tackle this problem more abstractly, introducing a categorical homotopy framework for LLMs. We introduce an LLM Markov category to represent probability distributions in language generated by an LLM, where the probability of a sentence, such as ``Charles Darwin wrote" is defined by an arrow in a Markov category. However, this approach runs into difficulties as language is full of equivalent rephrases, and each generates a non-isomorphic arrow in the LLM Markov category. To address this fundamental problem, we use categorical homotopy techniques to capture ``weak equivalences" in an LLM Markov category. We present a detailed overview of application of categorical homotopy to LLMs, from higher algebraic K-theory to model categories, building on powerful theoretical results developed over the past half a century.
|
[
{
"version": "v1",
"created": "Thu, 07 Aug 2025 00:48:30 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Mahadevan",
"Sridhar",
""
]
] |
2508.10019
|
Li Wang
|
Li Wang, Changhao Zhang, Zengqi Xiu, Kai Lu, Xin Yu, Kui Zhang, Wenjun Wu
|
Decoupling Understanding from Reasoning via Problem Space Mapping for Small-scale Model Reasoning
| null | null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Despite recent advances in the reasoning capabilities of Large Language Models (LLMs), improving the reasoning ability of Small Language Models (SLMs, e.g., $\leq$ 1.5B) remains challenging. A key obstacle lies in the complexity and variability of natural language: essentially equivalent problems often appear in diverse surface forms, often obscured by redundant or distracting details. This imposes a dual burden on SLMs: they must first extract the core problem from complex linguistic input, and then perform reasoning based on that understanding. The resulting vast and noisy problem space hinders optimization, particularly for models with limited capacity. To address this, we propose a new framework that decouples understanding from reasoning by mapping natural language problems into a canonical problem space-a semantically simplified yet expressive domain. This enables SLMs to focus on reasoning over standardized inputs, free from linguistic variability. Within this framework, we introduce DURIT (Decoupled Understanding from Reasoning via Iterative Training), a three-step algorithm that iteratively: (1) mapping natural language problems via reinforcement learning, (2) aligns reasoning trajectories through self-distillation, and (3) trains reasoning policies in the problem space. The mapper and reasoner are co-trained in an alternating loop throughout this process. Experiments show that DURIT substantially improves SLMs' performance on both in-domain and out-of-domain mathematical and logical reasoning tasks. Beyond improving reasoning capabilities, DURIT also improves the robustness of reasoning, validating decoupling understanding from reasoning as an effective strategy for strengthening SLMs.
|
[
{
"version": "v1",
"created": "Thu, 07 Aug 2025 01:13:30 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Wang",
"Li",
""
],
[
"Zhang",
"Changhao",
""
],
[
"Xiu",
"Zengqi",
""
],
[
"Lu",
"Kai",
""
],
[
"Yu",
"Xin",
""
],
[
"Zhang",
"Kui",
""
],
[
"Wu",
"Wenjun",
""
]
] |
2508.10020
|
Chuan Li
|
Chuan Li, Qianyi Zhao, Fengran Mo, Cen Chen
|
FedCoT: Communication-Efficient Federated Reasoning Enhancement for Large Language Models
| null | null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Efficiently enhancing the reasoning capabilities of large language models (LLMs) in federated learning environments remains challenging, particularly when balancing performance gains with strict computational, communication, and privacy constraints. This challenge is especially acute in healthcare, where decisions-spanning clinical, operational, and patient-facing contexts-demand not only accurate outputs but also interpretable, traceable rationales to ensure safety, accountability, and regulatory compliance. Conventional federated tuning approaches on LLM fail to address this need: they optimize primarily for answer correctness while neglecting rationale quality, leaving CoT capabilities dependent on models' innate pre-training abilities. Moreover, existing methods for improving rationales typically rely on privacy-violating knowledge distillation from centralized models. Additionally, the communication overhead in traditional federated fine-tuning on LLMs remains substantial. We addresses this gap by proposing FedCoT, a novel framework specifically designed to enhance reasoning in federated settings. FedCoT leverages a lightweight chain-of-thought enhancement mechanism: local models generate multiple reasoning paths, and a compact discriminator dynamically selects the most promising one. This approach improves reasoning accuracy and robustness while providing valuable interpretability, which is particularly critical for medical applications. To manage client heterogeneity efficiently, we adopt an improved aggregation approach building upon advanced LoRA module stacking, incorporating client classifier-awareness to achieve noise-free aggregation across diverse clients. Comprehensive experiments on medical reasoning tasks demonstrate that FedCoT significantly boosts client-side reasoning performance under stringent resource budgets while fully preserving data privacy.
|
[
{
"version": "v1",
"created": "Thu, 07 Aug 2025 06:50:15 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Li",
"Chuan",
""
],
[
"Zhao",
"Qianyi",
""
],
[
"Mo",
"Fengran",
""
],
[
"Chen",
"Cen",
""
]
] |
2508.10021
|
Egor Fadeev
|
Egor Fadeev, Dzhambulat Mollaev, Aleksei Shestov, Dima Korolev, Omar Zoloev, Ivan Kireev, Andrey Savchenko, Maksim Makarenko
|
LATTE: Learning Aligned Transactions and Textual Embeddings for Bank Clients
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Learning clients embeddings from sequences of their historic communications is central to financial applications. While large language models (LLMs) offer general world knowledge, their direct use on long event sequences is computationally expensive and impractical in real-world pipelines. In this paper, we propose LATTE, a contrastive learning framework that aligns raw event embeddings with semantic embeddings from frozen LLMs. Behavioral features are summarized into short prompts, embedded by the LLM, and used as supervision via contrastive loss. The proposed approach significantly reduces inference cost and input size compared to conventional processing of complete sequence by LLM. We experimentally show that our method outperforms state-of-the-art techniques for learning event sequence representations on real-world financial datasets while remaining deployable in latency-sensitive environments.
|
[
{
"version": "v1",
"created": "Thu, 07 Aug 2025 16:46:38 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Fadeev",
"Egor",
""
],
[
"Mollaev",
"Dzhambulat",
""
],
[
"Shestov",
"Aleksei",
""
],
[
"Korolev",
"Dima",
""
],
[
"Zoloev",
"Omar",
""
],
[
"Kireev",
"Ivan",
""
],
[
"Savchenko",
"Andrey",
""
],
[
"Makarenko",
"Maksim",
""
]
] |
2508.10022
|
YuanChang Ye
|
Yuanchang Ye
|
Conformal P-Value in Multiple-Choice Question Answering Tasks with Provable Risk Control
| null | null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This study introduces a significance testing-enhanced conformal prediction (CP) framework to improve trustworthiness of large language models (LLMs) in multiple-choice question answering (MCQA). While LLMs have been increasingly deployed in disciplinary QA scenarios, hallucination and nonfactual generation substantially compromise response reliability. Although CP provides statistically rigorous marginal coverage guarantees for prediction sets, and significance testing offers established statistical rigor, their synergistic integration remains unexplored. To mitigate hallucination and factual inaccuracies, our framework integrates $p$-value computation with conformity scoring through self-consistency resampling of MCQA responses. This approach calculates option frequencies to address LLMs' black-box nature, subsequently constructing prediction sets via null hypothesis testing ($\mathcal{H}_0$) with empirically derived $p$-values. Evaluations on MMLU and MMLU-Pro benchmarks using off-the-shelf LLMs demonstrate: (1) The enhanced CP achieves user-specified empirical miscoverage rates; (2) Test-set average prediction set size (APSS) decreases monotonically with increasing risk levels ($\alpha$), validating APSS as an effective uncertainty metric. This work establishes a principled statistical framework for trustworthy LLM deployment in high-stakes QA applications.
|
[
{
"version": "v1",
"created": "Thu, 07 Aug 2025 16:46:47 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Ye",
"Yuanchang",
""
]
] |
2508.10023
|
Samet \"Unsal
|
Samet \"Unsal
|
A Comparative Performance Evaluation of Kyber, sntrup761, and FrodoKEM for Post-Quantum Cryptography
|
12 pages, 3 tables, IEEE conference format
| null | null | null |
cs.CR quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Post-quantum cryptography (PQC) aims to develop cryptographic algorithms that are secure against attacks from quantum computers. This paper compares the leading postquantum cryptographic algorithms, such as Kyber, sntrup761, and FrodoKEM, in terms of their security, performance, and real-world applicability. The review highlights the strengths and weaknesses of each algorithm and provides insights into future research directions. We also discuss the challenges of transitioning from classical to post-quantum systems and the potential impacts on various industries. This paper serves as a foundation for understanding the current state of post-quantum cryptography and its future prospects in the quantum computing era.
|
[
{
"version": "v1",
"created": "Thu, 07 Aug 2025 19:39:59 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Ünsal",
"Samet",
""
]
] |
2508.10024
|
Juan Pablo Mu\~noz
|
J. Pablo Mu\~noz and Jinjie Yuan
|
RTTC: Reward-Guided Collaborative Test-Time Compute
| null | null | null | null |
cs.CL cs.AI cs.IR
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Test-Time Compute (TTC) has emerged as a powerful paradigm for enhancing the performance of Large Language Models (LLMs) at inference, leveraging strategies such as Test-Time Training (TTT) and Retrieval-Augmented Generation (RAG). However, the optimal adaptation strategy varies across queries, and indiscriminate application of TTC strategy incurs substantial computational overhead. In this work, we introduce Reward-Guided Test-Time Compute (RTTC), a novel framework that adaptively selects the most effective TTC strategy for each query via a pretrained reward model, maximizing downstream accuracy across diverse domains and tasks. RTTC operates in a distributed server-client architecture, retrieving relevant samples from a remote knowledge base and applying RAG or lightweight fine-tuning on client devices only when necessary. To further mitigate redundant computation, we propose Query-State Caching, which enables the efficient reuse of historical query states at both retrieval and adaptation levels. Extensive experiments across multiple LLMs and benchmarks demonstrate that RTTC consistently achieves superior accuracy compared to vanilla RAG or TTT, validating the necessity of adaptive, reward-guided TTC selection and the potential of RTTC for scalable, high-performance language model adaptation.
|
[
{
"version": "v1",
"created": "Thu, 07 Aug 2025 21:18:52 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Muñoz",
"J. Pablo",
""
],
[
"Yuan",
"Jinjie",
""
]
] |
2508.10025
|
Silvia Garc\'ia-M\'endez
|
Silvia Garc\'ia-M\'endez, Francisco de Arriba-P\'erez
|
Detecting and explaining postpartum depression in real-time with generative artificial intelligence
| null | null |
10.1080/08839514.2025.2515063
| null |
cs.CL cs.AI cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Among the many challenges mothers undergo after childbirth, postpartum depression (PPD) is a severe condition that significantly impacts their mental and physical well-being. Consequently, the rapid detection of ppd and their associated risk factors is critical for in-time assessment and intervention through specialized prevention procedures. Accordingly, this work addresses the need to help practitioners make decisions with the latest technological advancements to enable real-time screening and treatment recommendations. Mainly, our work contributes to an intelligent PPD screening system that combines Natural Language Processing, Machine Learning (ML), and Large Language Models (LLMs) towards an affordable, real-time, and non-invasive free speech analysis. Moreover, it addresses the black box problem since the predictions are described to the end users thanks to the combination of LLMs with interpretable ml models (i.e., tree-based algorithms) using feature importance and natural language. The results obtained are 90 % on ppd detection for all evaluation metrics, outperforming the competing solutions in the literature. Ultimately, our solution contributes to the rapid detection of PPD and their associated risk factors, critical for in-time and proper assessment and intervention.
|
[
{
"version": "v1",
"created": "Fri, 08 Aug 2025 07:57:05 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"García-Méndez",
"Silvia",
""
],
[
"de Arriba-Pérez",
"Francisco",
""
]
] |
2508.10026
|
Kai Zhao
|
Kai Zhao, Yanjun Zhao, Jiaming Song, Shien He, Lusheng Zhang, Qiang Zhang, Tianjiao Li
|
SABER: Switchable and Balanced Training for Efficient LLM Reasoning
| null | null | null | null |
cs.CL cs.AI cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large language models (LLMs) empowered by chain-of-thought reasoning have achieved impressive accuracy on complex tasks but suffer from excessive inference costs and latency when applied uniformly to all problems. We propose SABER (Switchable and Balanced Training for Efficient LLM Reasoning), a reinforcement learning framework that endows LLMs with user-controllable, token-budgeted reasoning. SABER first profiles each training example's base-model thinking token usage and assigns it to one of the predefined budget tiers. During fine-tuning, the model is guided by system prompts and length-aware rewards to respect its assigned budget. In parallel, we incorporate no-think examples to ensure the model remains reliable even when explicit reasoning is turned off. SABER further supports four discrete inference modes - NoThink, FastThink, CoreThink, and DeepThink, enabling flexible trade-offs between latency and reasoning depth. Extensive evaluations on math reasoning (MATH, GSM8K), code generation (MBPP), and logical reasoning (LiveBench-Reasoning) demonstrate that SABER achieves high accuracy under tight budgets, graceful degradation, and effective cross-scale and cross-domain generalization. In particular, SABER-FastThink cuts reasoning length by 65.4% and yields a 3.6% accuracy gain compared with the base model on the MATH benchmark.
|
[
{
"version": "v1",
"created": "Fri, 08 Aug 2025 11:27:48 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zhao",
"Kai",
""
],
[
"Zhao",
"Yanjun",
""
],
[
"Song",
"Jiaming",
""
],
[
"He",
"Shien",
""
],
[
"Zhang",
"Lusheng",
""
],
[
"Zhang",
"Qiang",
""
],
[
"Li",
"Tianjiao",
""
]
] |
2508.10027
|
Yasaman Haghbin
|
Ali Zolnour, Hossein Azadmaleki, Yasaman Haghbin, Fatemeh Taherinezhad, Mohamad Javad Momeni Nezhad, Sina Rashidi, Masoud Khani, AmirSajjad Taleban, Samin Mahdizadeh Sani, Maryam Dadkhah, James M. Noble, Suzanne Bakken, Yadollah Yaghoobzadeh, Abdol-Hossein Vahabie, Masoud Rouhizadeh, Maryam Zolnoori
|
LLMCARE: Alzheimer's Detection via Transformer Models Enhanced by LLM-Generated Synthetic Data
| null | null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Alzheimer's disease and related dementias (ADRD) affect approximately five million older adults in the U.S., yet over half remain undiagnosed. Speech-based natural language processing (NLP) offers a promising, scalable approach to detect early cognitive decline through linguistic markers.
To develop and evaluate a screening pipeline that (i) fuses transformer embeddings with handcrafted linguistic features, (ii) tests data augmentation using synthetic speech generated by large language models (LLMs), and (iii) benchmarks unimodal and multimodal LLM classifiers for ADRD detection.
Transcripts from the DementiaBank "cookie-theft" task (n = 237) were used. Ten transformer models were evaluated under three fine-tuning strategies. A fusion model combined embeddings from the top-performing transformer with 110 lexical-derived linguistic features. Five LLMs (LLaMA-8B/70B, MedAlpaca-7B, Ministral-8B, GPT-4o) were fine-tuned to generate label-conditioned synthetic speech, which was used to augment training data. Three multimodal models (GPT-4o, Qwen-Omni, Phi-4) were tested for speech-text classification in zero-shot and fine-tuned settings.
The fusion model achieved F1 = 83.3 (AUC = 89.5), outperforming linguistic or transformer-only baselines. Augmenting training data with 2x MedAlpaca-7B synthetic speech increased F1 to 85.7. Fine-tuning significantly improved unimodal LLM classifiers (e.g., MedAlpaca: F1 = 47.3 -> 78.5 F1). Current multimodal models demonstrated lower performance (GPT-4o = 70.2 F1; Qwen = 66.0). Performance gains aligned with the distributional similarity between synthetic and real speech.
Integrating transformer embeddings with linguistic features enhances ADRD detection from speech. Clinically tuned LLMs effectively support both classification and data augmentation, while further advancement is needed in multimodal modeling.
|
[
{
"version": "v1",
"created": "Fri, 08 Aug 2025 13:44:55 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zolnour",
"Ali",
""
],
[
"Azadmaleki",
"Hossein",
""
],
[
"Haghbin",
"Yasaman",
""
],
[
"Taherinezhad",
"Fatemeh",
""
],
[
"Nezhad",
"Mohamad Javad Momeni",
""
],
[
"Rashidi",
"Sina",
""
],
[
"Khani",
"Masoud",
""
],
[
"Taleban",
"AmirSajjad",
""
],
[
"Sani",
"Samin Mahdizadeh",
""
],
[
"Dadkhah",
"Maryam",
""
],
[
"Noble",
"James M.",
""
],
[
"Bakken",
"Suzanne",
""
],
[
"Yaghoobzadeh",
"Yadollah",
""
],
[
"Vahabie",
"Abdol-Hossein",
""
],
[
"Rouhizadeh",
"Masoud",
""
],
[
"Zolnoori",
"Maryam",
""
]
] |
2508.10028
|
Xiao Fu
|
Xiao Fu, Hossein A. Rahmani, Bin Wu, Jerome Ramos, Emine Yilmaz, Aldo Lipani
|
PREF: Reference-Free Evaluation of Personalised Text Generation in LLMs
|
7 pages
| null | null | null |
cs.CL cs.AI cs.HC cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Personalised text generation is essential for user-centric information systems, yet most evaluation methods overlook the individuality of users. We introduce \textbf{PREF}, a \textbf{P}ersonalised \textbf{R}eference-free \textbf{E}valuation \textbf{F}ramework that jointly measures general output quality and user-specific alignment without requiring gold personalised references. PREF operates in a three-step pipeline: (1) a coverage stage uses a large language model (LLM) to generate a comprehensive, query-specific guideline covering universal criteria such as factuality, coherence, and completeness; (2) a preference stage re-ranks and selectively augments these factors using the target user's profile, stated or inferred preferences, and context, producing a personalised evaluation rubric; and (3) a scoring stage applies an LLM judge to rate candidate answers against this rubric, ensuring baseline adequacy while capturing subjective priorities. This separation of coverage from preference improves robustness, transparency, and reusability, and allows smaller models to approximate the personalised quality of larger ones. Experiments on the PrefEval benchmark, including implicit preference-following tasks, show that PREF achieves higher accuracy, better calibration, and closer alignment with human judgments than strong baselines. By enabling scalable, interpretable, and user-aligned evaluation, PREF lays the groundwork for more reliable assessment and development of personalised language generation systems.
|
[
{
"version": "v1",
"created": "Fri, 08 Aug 2025 14:32:31 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Fu",
"Xiao",
""
],
[
"Rahmani",
"Hossein A.",
""
],
[
"Wu",
"Bin",
""
],
[
"Ramos",
"Jerome",
""
],
[
"Yilmaz",
"Emine",
""
],
[
"Lipani",
"Aldo",
""
]
] |
2508.10029
|
Wenpeng Xing
|
Wenpeng Xing, Mohan Li, Chunqiang Hu, Haitao XuNingyu Zhang, Bo Lin, Meng Han
|
Latent Fusion Jailbreak: Blending Harmful and Harmless Representations to Elicit Unsafe LLM Outputs
| null | null | null | null |
cs.CL cs.AI cs.CR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large language models (LLMs) demonstrate impressive capabilities in various language tasks but are susceptible to jailbreak attacks that circumvent their safety alignments. This paper introduces Latent Fusion Jailbreak (LFJ), a representation-based attack that interpolates hidden states from harmful and benign query pairs to elicit prohibited responses. LFJ begins by selecting query pairs with high thematic and syntactic similarity, then performs gradient-guided interpolation at influential layers and tokens, followed by optimization to balance attack success, output fluency, and computational efficiency. Evaluations on models such as Vicuna and LLaMA-2 across benchmarks like AdvBench and MaliciousInstruct yield an average attack success rate (ASR) of 94.01%, outperforming existing methods. To mitigate LFJ, we propose an adversarial training defense that fine-tunes models on interpolated examples, reducing ASR by over 80% without degrading performance on benign inputs. Ablation studies validate the importance of query pair selection, hidden state interpolation components, and optimization strategies in LFJ's effectiveness.
|
[
{
"version": "v1",
"created": "Fri, 08 Aug 2025 17:29:16 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Xing",
"Wenpeng",
""
],
[
"Li",
"Mohan",
""
],
[
"Hu",
"Chunqiang",
""
],
[
"Zhang",
"Haitao XuNingyu",
""
],
[
"Lin",
"Bo",
""
],
[
"Han",
"Meng",
""
]
] |
2508.10030
|
Saaduddin Mahmud
|
Saaduddin Mahmud, Mason Nakamura, Kyle H. Wray, Shlomo Zilberstein
|
Inference-Aware Prompt Optimization for Aligning Black-Box Large Language Models
|
17 pages
| null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Prompt optimization methods have demonstrated significant effectiveness in aligning black-box large language models (LLMs). In parallel, inference scaling strategies such as Best-of-N Sampling and Majority Voting have also proven to enhance alignment and performance by trading off computation. However, existing prompt optimization approaches are inference strategy agnostic; that is, they optimize prompts without regard to the inference strategy employed during deployment. This constitutes a significant methodological gap, as our empirical and theoretical analysis reveals a strong interdependence between these two paradigms. Moreover, we find that user preferences regarding trade-offs among multiple objectives and inference budgets substantially influence the choice of prompt and inference configuration. To address this gap, we introduce a unified novel framework named IAPO (Inference-Aware Prompt Optimization) that jointly optimizes the prompt and inference scale, while being aware of the inference budget and different task objectives. We then develop a fixed-budget training algorithm for IAPO, which we call PSST (Prompt Scaling via Sequential Trimming), and analyze finite-budget guarantees on error probability. Finally, we evaluate the effectiveness of PSST on six different tasks, including multi-objective text generation and reasoning, and demonstrate the critical role of incorporating inference-awareness when aligning black-box LLMs through prompt optimization.
|
[
{
"version": "v1",
"created": "Fri, 08 Aug 2025 18:45:53 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Mahmud",
"Saaduddin",
""
],
[
"Nakamura",
"Mason",
""
],
[
"Wray",
"Kyle H.",
""
],
[
"Zilberstein",
"Shlomo",
""
]
] |
2508.10031
|
Jinhwa Kim
|
Jinhwa Kim, Ian G. Harris
|
Context Misleads LLMs: The Role of Context Filtering in Maintaining Safe Alignment of LLMs
|
13 pages, 2 figures
| null | null | null |
cs.CR cs.AI cs.CL
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
While Large Language Models (LLMs) have shown significant advancements in performance, various jailbreak attacks have posed growing safety and ethical risks. Malicious users often exploit adversarial context to deceive LLMs, prompting them to generate responses to harmful queries. In this study, we propose a new defense mechanism called Context Filtering model, an input pre-processing method designed to filter out untrustworthy and unreliable context while identifying the primary prompts containing the real user intent to uncover concealed malicious intent. Given that enhancing the safety of LLMs often compromises their helpfulness, potentially affecting the experience of benign users, our method aims to improve the safety of the LLMs while preserving their original performance. We evaluate the effectiveness of our model in defending against jailbreak attacks through comparative analysis, comparing our approach with state-of-the-art defense mechanisms against six different attacks and assessing the helpfulness of LLMs under these defenses. Our model demonstrates its ability to reduce the Attack Success Rates of jailbreak attacks by up to 88% while maintaining the original LLMs' performance, achieving state-of-the-art Safety and Helpfulness Product results. Notably, our model is a plug-and-play method that can be applied to all LLMs, including both white-box and black-box models, to enhance their safety without requiring any fine-tuning of the models themselves. We will make our model publicly available for research purposes.
|
[
{
"version": "v1",
"created": "Sat, 09 Aug 2025 02:37:59 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Kim",
"Jinhwa",
""
],
[
"Harris",
"Ian G.",
""
]
] |
2508.10032
|
Fan Yang
|
Fan Yang
|
The Cost of Thinking: Increased Jailbreak Risk in Large Language Models
| null | null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Thinking mode has always been regarded as one of the most valuable modes in LLMs. However, we uncover a surprising and previously overlooked phenomenon: LLMs with thinking mode are more easily broken by Jailbreak attack. We evaluate 9 LLMs on AdvBench and HarmBench and find that the success rate of attacking thinking mode in LLMs is almost higher than that of non-thinking mode. Through large numbers of sample studies, it is found that for educational purposes and excessively long thinking lengths are the characteristics of successfully attacked data, and LLMs also give harmful answers when they mostly know that the questions are harmful. In order to alleviate the above problems, this paper proposes a method of safe thinking intervention for LLMs, which explicitly guides the internal thinking processes of LLMs by adding "specific thinking tokens" of LLMs to the prompt. The results demonstrate that the safe thinking intervention can significantly reduce the attack success rate of LLMs with thinking mode.
|
[
{
"version": "v1",
"created": "Sat, 09 Aug 2025 09:49:49 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Yang",
"Fan",
""
]
] |
2508.10033
|
Yuksel Aydin
|
Yuksel Aydin
|
Cognitive Cybersecurity for Artificial Intelligence: Guardrail Engineering with CCS-7
| null | null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Language models exhibit human-like cognitive vulnerabilities, such as emotional framing, that escape traditional behavioral alignment. We present CCS-7 (Cognitive Cybersecurity Suite), a taxonomy of seven vulnerabilities grounded in human cognitive security research. To establish a human benchmark, we ran a randomized controlled trial with 151 participants: a "Think First, Verify Always" (TFVA) lesson improved cognitive security by +7.9% overall. We then evaluated TFVA-style guardrails across 12,180 experiments on seven diverse language model architectures. Results reveal architecture-dependent risk patterns: some vulnerabilities (e.g., identity confusion) are almost fully mitigated, while others (e.g., source interference) exhibit escalating backfire, with error rates increasing by up to 135% in certain models. Humans, in contrast, show consistent moderate improvement. These findings reframe cognitive safety as a model-specific engineering problem: interventions effective in one architecture may fail, or actively harm, another, underscoring the need for architecture-aware cognitive safety testing before deployment.
|
[
{
"version": "v1",
"created": "Sat, 09 Aug 2025 15:46:30 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Aydin",
"Yuksel",
""
]
] |
2508.10034
|
Arghya Chattopadhyay
|
Juvenal Bassa, Vidya Manian, Sudhir Malik, Arghya Chattopadhyay
|
Jet Image Tagging Using Deep Learning: An Ensemble Model
|
19 Pages. All codes available at https://github.com/Basjuven/Jet_Images_Tagging_EM
| null | null | null |
physics.data-an cs.AI cs.LG hep-ex hep-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Jet classification in high-energy particle physics is important for understanding fundamental interactions and probing phenomena beyond the Standard Model. Jets originate from the fragmentation and hadronization of quarks and gluons, and pose a challenge for identification due to their complex, multidimensional structure. Traditional classification methods often fall short in capturing these intricacies, necessitating advanced machine learning approaches. In this paper, we employ two neural networks simultaneously as an ensemble to tag various jet types. We convert the jet data to two-dimensional histograms instead of representing them as points in a higher-dimensional space. Specifically, this ensemble approach, hereafter referred to as Ensemble Model, is used to tag jets into classes from the JetNet dataset, corresponding to: Top Quarks, Light Quarks (up or down), and W and Z bosons. For the jet classes mentioned above, we show that the Ensemble Model can be used for both binary and multi-categorical classification. This ensemble approach learns jet features by leveraging the strengths of each constituent network achieving superior performance compared to either individual network.
|
[
{
"version": "v1",
"created": "Sat, 09 Aug 2025 17:40:15 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Bassa",
"Juvenal",
""
],
[
"Manian",
"Vidya",
""
],
[
"Malik",
"Sudhir",
""
],
[
"Chattopadhyay",
"Arghya",
""
]
] |
2508.10035
|
Biswash Basnet Basnet
|
Varsha Sen, Biswash Basnet
|
Neural Network-Based Detection and Multi-Class Classification of FDI Attacks in Smart Grid Home Energy Systems
|
17 pages, 7 figures
| null | null | null |
cs.CR cs.LG cs.SY eess.SY
|
http://creativecommons.org/publicdomain/zero/1.0/
|
False Data Injection Attacks (FDIAs) pose a significant threat to smart grid infrastructures, particularly Home Area Networks (HANs), where real-time monitoring and control are highly adopted. Owing to the comparatively less stringent security controls and widespread availability of HANs, attackers view them as an attractive entry point to manipulate aggregated demand patterns, which can ultimately propagate and disrupt broader grid operations. These attacks undermine the integrity of smart meter data, enabling malicious actors to manipulate consumption values without activating conventional alarms, thereby creating serious vulnerabilities across both residential and utility-scale infrastructures. This paper presents a machine learning-based framework for both the detection and classification of FDIAs using residential energy data. A real-time detection is provided by the lightweight Artificial Neural Network (ANN), which works by using the most vital features of energy consumption, cost, and time context. For the classification of different attack types, a Bidirectional LSTM is trained to recognize normal, trapezoidal, and sigmoid attack shapes through learning sequential dependencies in the data. A synthetic time-series dataset was generated to emulate realistic household behaviour. Experimental results demonstrate that the proposed models are effective in identifying and classifying FDIAs, offering a scalable solution for enhancing grid resilience at the edge. This work contributes toward building intelligent, data-driven defence mechanisms that strengthen smart grid cybersecurity from residential endpoints.
|
[
{
"version": "v1",
"created": "Sat, 09 Aug 2025 18:27:31 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Sen",
"Varsha",
""
],
[
"Basnet",
"Biswash",
""
]
] |
2508.10036
|
Xiang Chen
|
Dong Zhao, Yadong Wang, Xiang Chen, Chenxi Wang, Hongliang Dai, Chuanxing Geng, Shengzhong Zhang, Shaoyuan Li, Sheng-Jun Huang
|
Reflect then Learn: Active Prompting for Information Extraction Guided by Introspective Confusion
|
Under Review
| null | null | null |
cs.CL cs.AI cs.IR cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large Language Models (LLMs) show remarkable potential for few-shot information extraction (IE), yet their performance is highly sensitive to the choice of in-context examples. Conventional selection strategies often fail to provide informative guidance, as they overlook a key source of model fallibility: confusion stemming not just from semantic content, but also from the generation of well-structured formats required by IE tasks. To address this, we introduce Active Prompting for Information Extraction (APIE), a novel active prompting framework guided by a principle we term introspective confusion. Our method empowers an LLM to assess its own confusion through a dual-component uncertainty metric that uniquely quantifies both Format Uncertainty (difficulty in generating correct syntax) and Content Uncertainty (inconsistency in extracted semantics). By ranking unlabeled data with this comprehensive score, our framework actively selects the most challenging and informative samples to serve as few-shot exemplars. Extensive experiments on four benchmarks show that our approach consistently outperforms strong baselines, yielding significant improvements in both extraction accuracy and robustness. Our work highlights the critical importance of a fine-grained, dual-level view of model uncertainty when it comes to building effective and reliable structured generation systems.
|
[
{
"version": "v1",
"created": "Sun, 10 Aug 2025 02:27:41 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zhao",
"Dong",
""
],
[
"Wang",
"Yadong",
""
],
[
"Chen",
"Xiang",
""
],
[
"Wang",
"Chenxi",
""
],
[
"Dai",
"Hongliang",
""
],
[
"Geng",
"Chuanxing",
""
],
[
"Zhang",
"Shengzhong",
""
],
[
"Li",
"Shaoyuan",
""
],
[
"Huang",
"Sheng-Jun",
""
]
] |
2508.10037
|
Xiufang Chen
|
Jiang Peng and Xi Liu and Zhenhua Chai and Changsheng Huang and Xiufang Chen
|
Electrohydrodynamics of a pair of leaky dielectric droplets on the solid substrate: A lattice Boltzmann study
| null | null | null | null |
physics.flu-dyn
|
http://creativecommons.org/licenses/by/4.0/
|
In this work, the electrohydrodynamics of a pair of leaky dielectric droplets on a solid substrate is investigated by the phase-field-based lattice Boltzmann method. Different from a pair of suspended droplets that may coalesce or separate, two leaky dielectric droplets on the substrate exhibit more complex modes due to the effects of wettability and electric force. The results show that when a horizontal electric field is applied, five different modes with electrostatic attractive force are observed, including attraction without coalescence, attraction with coalescence, coalescence with bubble entrapment, coalescence followed by suspension, and suspension followed by coalescence. Particularly, if the droplets are in a hydrophilic state, the coalescence mode is usually observed, while for droplets in a neutral or hydrophobic state, the permittivity ratio has an important effect on the droplet modes. Additionally, during the coalescence process, two droplets in a hydrophobic state not only capture bubbles, but may also exhibit suspension at a large permittivity ratio or contact angle. On the other hand, when a vertical electric field is applied, there are three different modes with repulsive electrostatic force, including non-coalescence, coalescence, and suspension followed by repulsion. Specially, a small permittivity ratio or a large contact angle can suppress the horizontal deformation of droplets, preventing their coalescence. Moreover, under superhydrophobic conditions, both horizontal and vertical electric fields suspend the droplets. However, the vertical electric field induces repulsion between the suspended droplets, driving them apart, whereas the horizontal electric field promotes their coalescence.
|
[
{
"version": "v1",
"created": "Sun, 10 Aug 2025 05:33:22 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Peng",
"Jiang",
""
],
[
"Liu",
"Xi",
""
],
[
"Chai",
"Zhenhua",
""
],
[
"Huang",
"Changsheng",
""
],
[
"Chen",
"Xiufang",
""
]
] |
2508.10038
|
Sarath Sivaprasad
|
Pierre-Francois Gimenez, Sarath Sivaprasad, Mario Fritz
|
Certifiably robust malware detectors by design
| null | null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/publicdomain/zero/1.0/
|
Malware analysis involves analyzing suspicious software to detect malicious payloads. Static malware analysis, which does not require software execution, relies increasingly on machine learning techniques to achieve scalability. Although such techniques obtain very high detection accuracy, they can be easily evaded with adversarial examples where a few modifications of the sample can dupe the detector without modifying the behavior of the software. Unlike other domains, such as computer vision, creating an adversarial example of malware without altering its functionality requires specific transformations. We propose a new model architecture for certifiably robust malware detection by design. In addition, we show that every robust detector can be decomposed into a specific structure, which can be applied to learn empirically robust malware detectors, even on fragile features. Our framework ERDALT is based on this structure. We compare and validate these approaches with machine-learning-based malware detection methods, allowing for robust detection with limited reduction of detection performance.
|
[
{
"version": "v1",
"created": "Sun, 10 Aug 2025 09:19:29 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Gimenez",
"Pierre-Francois",
""
],
[
"Sivaprasad",
"Sarath",
""
],
[
"Fritz",
"Mario",
""
]
] |
2508.10039
|
Yangshijie Zhang
|
Wenqiang Wang, Yan Xiao, Hao Lin, Yangshijie Zhang, Xiaochun Cao
|
Multi-task Adversarial Attacks against Black-box Model with Few-shot Queries
| null | null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Current multi-task adversarial text attacks rely on abundant access to shared internal features and numerous queries, often limited to a single task type. As a result, these attacks are less effective against practical scenarios involving black-box feedback APIs, limited queries, or multiple task types. To bridge this gap, we propose \textbf{C}luster and \textbf{E}nsemble \textbf{M}ulti-task Text Adversarial \textbf{A}ttack (\textbf{CEMA}), an effective black-box attack that exploits the transferability of adversarial texts across different tasks. CEMA simplifies complex multi-task scenarios by using a \textit{deep-level substitute model} trained in a \textit{plug-and-play} manner for text classification, enabling attacks without mimicking the victim model. This approach requires only a few queries for training, converting multi-task attacks into classification attacks and allowing attacks across various tasks.
CEMA generates multiple adversarial candidates using different text classification methods and selects the one that most effectively attacks substitute models.
In experiments involving multi-task models with two, three, or six tasks--spanning classification, translation, summarization, and text-to-image generation--CEMA demonstrates significant attack success with as few as 100 queries. Furthermore, CEMA can target commercial APIs (e.g., Baidu and Google Translate), large language models (e.g., ChatGPT 4o), and image-generation models (e.g., Stable Diffusion V2), showcasing its versatility and effectiveness in real-world applications.
|
[
{
"version": "v1",
"created": "Sun, 10 Aug 2025 12:46:47 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Wang",
"Wenqiang",
""
],
[
"Xiao",
"Yan",
""
],
[
"Lin",
"Hao",
""
],
[
"Zhang",
"Yangshijie",
""
],
[
"Cao",
"Xiaochun",
""
]
] |
2508.10040
|
V\'itor Louren\c{c}o
|
V\'itor N. Louren\c{c}o and Aline Paes and and Tillman Weyde
|
Exploring Content and Social Connections of Fake News with Explainable Text and Graph Learning
|
Accepted to publication at the 35th Brazilian Conference on Intelligent Systems, BRACIS 2025
| null | null | null |
cs.SI cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
The global spread of misinformation and concerns about content trustworthiness have driven the development of automated fact-checking systems. Since false information often exploits social media dynamics such as "likes" and user networks to amplify its reach, effective solutions must go beyond content analysis to incorporate these factors. Moreover, simply labelling content as false can be ineffective or even reinforce biases such as automation and confirmation bias. This paper proposes an explainable framework that combines content, social media, and graph-based features to enhance fact-checking. It integrates a misinformation classifier with explainability techniques to deliver complete and interpretable insights supporting classification decisions. Experiments demonstrate that multimodal information improves performance over single modalities, with evaluations conducted on datasets in English, Spanish, and Portuguese. Additionally, the framework's explanations were assessed for interpretability, trustworthiness, and robustness with a novel protocol, showing that it effectively generates human-understandable justifications for its predictions.
|
[
{
"version": "v1",
"created": "Mon, 11 Aug 2025 12:03:37 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Lourenço",
"Vítor N.",
""
],
[
"Paes",
"Aline",
""
],
[
"Weyde",
"and Tillman",
""
]
] |
2508.10041
|
Julien Mellaerts
|
Julien Mellaerts
|
Quantum Prime Factorization: A Novel Approach Based on Fermat Method
| null | null | null | null |
cs.CR quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
In this paper, we introduce a novel quantum algorithm for the factorization of composite odd numbers. This work makes two significant contributions. First, we present a new improvement to the classical Fermat method, fourfold reducing the computational complexity of factoring. Second, we reformulate Fermat factorization method as an optimization problem suitable for Quantum Annealers which allowed us to factorize 8,689,739, the biggest number ever factorized using a quantum device to our knowledge.
|
[
{
"version": "v1",
"created": "Mon, 11 Aug 2025 13:46:23 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Mellaerts",
"Julien",
""
]
] |
2508.10042
|
Jane Carney
|
Jane Carney, Kushal Upreti, Gaby G. Dagher, Tim Andersen
|
FIDELIS: Blockchain-Enabled Protection Against Poisoning Attacks in Federated Learning
| null | null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Federated learning enhances traditional deep learning by enabling the joint training of a model with the use of IoT device's private data. It ensures privacy for clients, but is susceptible to data poisoning attacks during training that degrade model performance and integrity. Current poisoning detection methods in federated learning lack a standardized detection method or take significant liberties with trust. In this paper, we present \Sys, a novel blockchain-enabled poison detection framework in federated learning. The framework decentralizes the role of the global server across participating clients. We introduce a judge model used to detect data poisoning in model updates. The judge model is produced by each client and verified to reach consensus on a single judge model. We implement our solution to show \Sys is robust against data poisoning attacks and the creation of our judge model is scalable.
|
[
{
"version": "v1",
"created": "Mon, 11 Aug 2025 22:12:27 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Carney",
"Jane",
""
],
[
"Upreti",
"Kushal",
""
],
[
"Dagher",
"Gaby G.",
""
],
[
"Andersen",
"Tim",
""
]
] |
2508.10043
|
Pallavi Zambare
|
Pallavi Zambare, Venkata Nikhil Thanikella, Ying Liu
|
Securing Agentic AI: Threat Modeling and Risk Analysis for Network Monitoring Agentic AI System
|
Submitted and under review in IEEE Transactions on Privacy
| null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
When combining Large Language Models (LLMs) with autonomous agents, used in network monitoring and decision-making systems, this will create serious security issues. In this research, the MAESTRO framework consisting of the seven layers threat modeling architecture in the system was used to expose, evaluate, and eliminate vulnerabilities of agentic AI. The prototype agent system was constructed and implemented, using Python, LangChain, and telemetry in WebSockets, and deployed with inference, memory, parameter tuning, and anomaly detection modules. Two practical threat cases were confirmed as follows: (i) resource denial of service by traffic replay denial-of-service, and (ii) memory poisoning by tampering with the historical log file maintained by the agent. These situations resulted in measurable levels of performance degradation, i.e. telemetry updates were delayed, and computational loads were increased, as a result of poor system adaptations. It was suggested to use a multilayered defense-in-depth approach with memory isolation, validation of planners and anomaly response systems in real-time. These findings verify that MAESTRO is viable in operational threat mapping, prospective risk scoring, and the basis of the resilient system design. The authors bring attention to the importance of the enforcement of memory integrity, paying attention to the adaptation logic monitoring, and cross-layer communication protection that guarantee the agentic AI reliability in adversarial settings.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 00:14:12 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zambare",
"Pallavi",
""
],
[
"Thanikella",
"Venkata Nikhil",
""
],
[
"Liu",
"Ying",
""
]
] |
2508.10044
|
Aydin Zaboli
|
Aydin Zaboli and Junho Hong
|
Generative AI for Cybersecurity of Energy Management Systems: Methods, Challenges, and Future Directions
|
36 pages, 10 figures
| null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
This paper elaborates on an extensive security framework specifically designed for energy management systems (EMSs), which effectively tackles the dynamic environment of cybersecurity vulnerabilities and/or system problems (SPs), accomplished through the incorporation of novel methodologies. A comprehensive multi-point attack/error model is initially proposed to systematically identify vulnerabilities throughout the entire EMS data processing pipeline, including post state estimation (SE) stealth attacks, EMS database manipulation, and human-machine interface (HMI) display corruption according to the real-time database (RTDB) storage. This framework acknowledges the interconnected nature of modern attack vectors, which utilize various phases of supervisory control and data acquisition (SCADA) data flow. Then, generative AI (GenAI)-based anomaly detection systems (ADSs) for EMSs are proposed for the first time in the power system domain to handle the scenarios. Further, a set-of-mark generative intelligence (SoM-GI) framework, which leverages multimodal analysis by integrating visual markers with rules considering the GenAI capabilities, is suggested to overcome inherent spatial reasoning limitations. The SoM-GI methodology employs systematic visual indicators to enable accurate interpretation of segmented HMI displays and detect visual anomalies that numerical methods fail to identify. Validation on the IEEE 14-Bus system shows the framework's effectiveness across scenarios, while visual analysis identifies inconsistencies. This integrated approach combines numerical analysis with visual pattern recognition and linguistic rules to protect against cyber threats and system errors.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 03:10:22 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zaboli",
"Aydin",
""
],
[
"Hong",
"Junho",
""
]
] |
2508.10045
|
Mikail Rubinov
|
Mika Rubinov
|
Unifying equivalences across unsupervised learning, network science, and imaging/network neuroscience
| null | null | null | null |
q-bio.NC
|
http://creativecommons.org/licenses/by/4.0/
|
Modern scientific fields face the challenge of integrating a wealth of data, analyses, and results. We recently showed that a neglect of this integration can lead to circular analyses and redundant explanations. Here, we help advance scientific integration by describing equivalences that unify diverse analyses of datasets and networks. We describe equivalences across analyses of clustering and dimensionality reduction, network centrality and dynamics, and popular models in imaging and network neuroscience. First, we equate foundational objectives across unsupervised learning and network science (from k means to modularity to UMAP), fuse classic algorithms for optimizing these objectives, and extend these objectives to simplify interpretations of popular dimensionality reduction methods. Second, we equate basic measures of connectional magnitude and dispersion with six measures of communication, control, and diversity in network science and network neuroscience. Third, we describe three semi-analytical vignettes that clarify and simplify the interpretation of structural and dynamical analyses in imaging and network neuroscience. We illustrate our results on example brain-imaging data and provide abct, an open multi-language toolbox that implements our analyses. Together, our study unifies diverse analyses across unsupervised learning, network science, imaging neuroscience, and network neuroscience.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 06:28:49 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Rubinov",
"Mika",
""
]
] |
2508.10046
|
Muhammad Ahmad
|
Muhammad Ahmad, Fida Ullah, Muhammad Usman, Ildar Batyrshin, Grigori Sidorov
|
SABIA: An AI-Powered Tool for Detecting Opioid-Related Behaviors on Social Media
| null | null | null | null |
cs.SI cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Social media platforms have become valuable tools for understanding public health challenges by offering insights into patient behaviors, medication use, and mental health issues. However, analyzing such data remains difficult due to the prevalence of informal language, slang, and coded communication, which can obscure the detection of opioid misuse. This study addresses the issue of opioid-related user behavior on social media, including informal expressions, slang terms, and misspelled or coded language. We analyzed the existing Bidirectional Encoder Representations from Transformers (BERT) technique and developed a BERT-BiLSTM-3CNN hybrid deep learning model, named SABIA, to create a single-task classifier that effectively captures the features of the target dataset. The SABIA model demonstrated strong capabilities in capturing semantics and contextual information. The proposed approach includes: (1) data preprocessing, (2) data representation using the SABIA model, (3) a fine-tuning phase, and (4) classification of user behavior into five categories. A new dataset was constructed from Reddit posts, identifying opioid user behaviors across five classes: Dealers, Active Opioid Users, Recovered Users, Prescription Users, and Non-Users, supported by detailed annotation guidelines. Experiments were conducted using supervised learning. Results show that SABIA achieved benchmark performance, outperforming the baseline (Logistic Regression, LR = 0.86) and improving accuracy by 9.30%. Comparisons with seven previous studies confirmed its effectiveness and robustness. This study demonstrates the potential of hybrid deep learning models for detecting complex opioid-related behaviors on social media, supporting public health monitoring and intervention efforts.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 06:52:41 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Ahmad",
"Muhammad",
""
],
[
"Ullah",
"Fida",
""
],
[
"Usman",
"Muhammad",
""
],
[
"Batyrshin",
"Ildar",
""
],
[
"Sidorov",
"Grigori",
""
]
] |
2508.10047
|
Ziyang Xiao
|
Ziyang Xiao, Jingrong Xie, Lilin Xu, Shisi Guan, Jingyan Zhu, Xiongwei Han, Xiaojin Fu, WingYin Yu, Han Wu, Wei Shi, Qingcan Kang, Jiahui Duan, Tao Zhong, Mingxuan Yuan, Jia Zeng, Yuan Wang, Gang Chen, Dongxiang Zhang
|
A Survey of Optimization Modeling Meets LLMs: Progress and Future Directions
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
By virtue of its great utility in solving real-world problems, optimization modeling has been widely employed for optimal decision-making across various sectors, but it requires substantial expertise from operations research professionals. With the advent of large language models (LLMs), new opportunities have emerged to automate the procedure of mathematical modeling. This survey presents a comprehensive and timely review of recent advancements that cover the entire technical stack, including data synthesis and fine-tuning for the base model, inference frameworks, benchmark datasets, and performance evaluation. In addition, we conducted an in-depth analysis on the quality of benchmark datasets, which was found to have a surprisingly high error rate. We cleaned the datasets and constructed a new leaderboard with fair performance evaluation in terms of base LLM model and datasets. We also build an online portal that integrates resources of cleaned datasets, code and paper repository to benefit the community. Finally, we identify limitations in current methodologies and outline future research opportunities.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 06:55:33 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Xiao",
"Ziyang",
""
],
[
"Xie",
"Jingrong",
""
],
[
"Xu",
"Lilin",
""
],
[
"Guan",
"Shisi",
""
],
[
"Zhu",
"Jingyan",
""
],
[
"Han",
"Xiongwei",
""
],
[
"Fu",
"Xiaojin",
""
],
[
"Yu",
"WingYin",
""
],
[
"Wu",
"Han",
""
],
[
"Shi",
"Wei",
""
],
[
"Kang",
"Qingcan",
""
],
[
"Duan",
"Jiahui",
""
],
[
"Zhong",
"Tao",
""
],
[
"Yuan",
"Mingxuan",
""
],
[
"Zeng",
"Jia",
""
],
[
"Wang",
"Yuan",
""
],
[
"Chen",
"Gang",
""
],
[
"Zhang",
"Dongxiang",
""
]
] |
2508.10048
|
Ville Virtanen
|
V.A. Virtanen, T. Eronen, A. Kankainen, O. Beliuskina, Z. Ge, R.P. de Groote, A. Jokinen, M. Mougeot, A. de Roubin, J. Ruotsalainen, J. Sar\'en, A. Takkinen
|
High-resolution multi-reflection time-of-flight mass spectrometer for exotic nuclei at IGISOL
| null | null | null | null |
physics.ins-det nucl-ex
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
A Multi-Reflection Time-of-Flight Mass Spectrometer (MR-ToF-MS) has been commissioned at the Ion-Guide Isotope Separator On-Line (IGISOL) facility. It consists of six electrode pairs that form a nearly energy-isochronous potential and a pulsed drift-tube to trap the ions between the electrodes. Time-of-flight peak widths down to 22 ns full-width at half-maximum and mass-resolving powers of $\approx 1.5 \times 10^5$ within 20 ms have been demonstrated. The obtained time-focus and mass-resolving power depend sensitively on the trapping energy, energy spread and the number of revolutions of the ions. The mass-resolving power is affected by the temporal and energy spread of the ions entering the MR-ToF-MS, and fluctuations in the electrode voltages due to temperature variations. The longitudinal emittance corresponding to the temporal and energy spread of $^{39}$K is estimated to be 175 eVns based on the data, close to the expected 186(10) eVns. The time-of-flight temperature sensitivity is found to be -5.55(30) ppm/K. In addition to atomic mass measurements of short-lived exotic nuclides, the MR-ToF-MS can be used as a fast mass separator for various other experiments at IGISOL and as an ion counter for laser spectroscopy and yield measurements.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 09:20:39 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Virtanen",
"V. A.",
""
],
[
"Eronen",
"T.",
""
],
[
"Kankainen",
"A.",
""
],
[
"Beliuskina",
"O.",
""
],
[
"Ge",
"Z.",
""
],
[
"de Groote",
"R. P.",
""
],
[
"Jokinen",
"A.",
""
],
[
"Mougeot",
"M.",
""
],
[
"de Roubin",
"A.",
""
],
[
"Ruotsalainen",
"J.",
""
],
[
"Sarén",
"J.",
""
],
[
"Takkinen",
"A.",
""
]
] |
2508.10049
|
Masahiro Morikawa
|
Akika Nakamichi and Izumi Uesaka and Masahiro Morikawa
|
Dynamic Synchronization and Resonance as a Universal Origin of 1/f Fluctuations -- Amplitude Modulation Across Music and Nature
|
14 pages, 10 figures
| null | null | null |
cs.SD
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose a universal physical mechanism for the emergence of 1/f fluctuations, observed across a wide range of systems. In particular, we verify this on acoustic cases. The mechanism is based on amplitude modulation (AM) and demodulation (DM), where the 1/f spectral law arises not in the raw waveform but in its demodulated amplitude envelope. Two distinct yet complementary processes generate the required AM: (i) stochastic synchronization among oscillators, modeled via an extended Kuramoto framework that captures perpetual synchronization-desynchronization cycles, and (ii) frequency-selective resonance, modeled by spectral accumulation of eigenmodes in acoustic or structural environments. Numerical simulations demonstrate that both mechanisms, acting separately or in combination, robustly produce 1/f spectra over several decades when DM is applied, and that the classical Kuramoto critical point is not necessary for their emergence. We demonstrate the cross-domain relevance of this AM/DM framework through analyses of musical performances, seismic records, and astrophysical time series, revealing a common underlying structure. This work establishes demodulation as a general route to 1/f fluctuations, providing a simple and scalable explanation for its ubiquity in both natural and engineered systems.
Keywords: 1/f fluctuation, amplitude modulation, synchronization, resonance, Kuramoto model, music, natural noise, demodulation
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 10:23:16 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Nakamichi",
"Akika",
""
],
[
"Uesaka",
"Izumi",
""
],
[
"Morikawa",
"Masahiro",
""
]
] |
2508.10050
|
Nathan Sherburn
|
Greg Sadler, Nathan Sherburn
|
Legal Zero-Days: A Novel Risk Vector for Advanced AI Systems
|
10 pages, 1 table, 1 figure. Introduces Legal Zero-Days as a novel AI risk vector and provides evaluation framework for measuring AI systems' ability to discover legal vulnerabilities
| null | null | null |
cs.CY cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
We introduce the concept of "Legal Zero-Days" as a novel risk vector for advanced AI systems. Legal Zero-Days are previously undiscovered vulnerabilities in legal frameworks that, when exploited, can cause immediate and significant societal disruption without requiring litigation or other processes before impact. We present a risk model for identifying and evaluating these vulnerabilities, demonstrating their potential to bypass safeguards or impede government responses to AI incidents. Using the 2017 Australian dual citizenship crisis as a case study, we illustrate how seemingly minor legal oversights can lead to large-scale governance disruption. We develop a methodology for creating "legal puzzles" as evaluation instruments for assessing AI systems' capabilities to discover such vulnerabilities. Our findings suggest that while current AI models may not reliably find impactful Legal Zero-Days, future systems may develop this capability, presenting both risks and opportunities for improving legal robustness. This work contributes to the broader effort to identify and mitigate previously unrecognized risks from frontier AI systems.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 11:43:00 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Sadler",
"Greg",
""
],
[
"Sherburn",
"Nathan",
""
]
] |
2508.10051
|
Satrajit S. Ghosh
|
Satrajit S. Ghosh
|
An Intelligent Infrastructure as a Foundation for Modern Science
| null | null | null | null |
q-bio.NC cs.ET
|
http://creativecommons.org/licenses/by/4.0/
|
Infrastructure shapes societies and scientific discovery. Traditional scientific infrastructure, often static and fragmented, leads to issues like data silos, lack of interoperability and reproducibility, and unsustainable short-lived solutions. Our current technical inability and social reticence to connect and coordinate scientific research and engineering leads to inefficiencies and impedes progress. With AI technologies changing how we interact with the world around us, there is an opportunity to transform scientific processes. Neuroscience's exponential growth of multimodal and multiscale data, and urgent clinical relevance demand an infrastructure itself learns, coordinates, and improves. Using neuroscience as a stress test, this perspective argues for a paradigm shift: infrastructure must evolve into a dynamic, AI-aligned ecosystem to accelerate science. Building on several existing principles for data, collective benefit, and digital repositories, I recommend operational guidelines for implementing them to create this dynamic ecosystem, aiming to foster a decentralized, self-learning, and self-correcting system where humans and AI can collaborate seamlessly. Addressing the chronic underfunding of scientific infrastructure, acknowledging diverse contributions beyond publications, and coordinating global efforts are critical steps for this transformation. By prioritizing an intelligent infrastructure as a central scientific instrument for knowledge generation, we can overcome current limitations, accelerate discovery, ensure reproducibility and ethical practices, and ultimately translate neuroscientific understanding into tangible societal benefits, setting a blueprint for other scientific domains.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 14:10:57 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Ghosh",
"Satrajit S.",
""
]
] |
2508.10052
|
Pallavi Zambare
|
Pallavi Zambare, Venkata Nikhil Thanikella, Nikhil Padmanabh Kottur, Sree Akhil Akula, Ying Liu
|
NetMoniAI: An Agentic AI Framework for Network Security & Monitoring
|
Accepted in IEEE 3rd International Conference on Artificial Intelligence, Blockchain, and Internet of Things (AIBThings 2025)
| null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
In this paper, we present NetMoniAI, an agentic AI framework for automatic network monitoring and security that integrates decentralized analysis with lightweight centralized coordination. The framework consists of two layers: autonomous micro-agents at each node perform local traffic analysis and anomaly detection. A central controller then aggregates insights across nodes to detect coordinated attacks and maintain system-wide situational awareness. We evaluated NetMoniAI on a local micro-testbed and through NS-3 simulations. Results confirm that the two-tier agentic-AI design scales under resource constraints, reduces redundancy, and improves response time without compromising accuracy. To facilitate broader adoption and reproducibility, the complete framework is available as open source. This enables researchers and practitioners to replicate, validate, and extend it across diverse network environments and threat scenarios. Github link: https://github.com/pzambare3/NetMoniAI
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 15:48:53 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zambare",
"Pallavi",
""
],
[
"Thanikella",
"Venkata Nikhil",
""
],
[
"Kottur",
"Nikhil Padmanabh",
""
],
[
"Akula",
"Sree Akhil",
""
],
[
"Liu",
"Ying",
""
]
] |
2508.10053
|
Daniel Beaglehole
|
Daniel Beaglehole, David Holzm\"uller, Adityanarayanan Radhakrishnan, Mikhail Belkin
|
xRFM: Accurate, scalable, and interpretable feature learning models for tabular data
| null | null | null | null |
cs.LG stat.ML
|
http://creativecommons.org/publicdomain/zero/1.0/
|
Inference from tabular data, collections of continuous and categorical variables organized into matrices, is a foundation for modern technology and science. Yet, in contrast to the explosive changes in the rest of AI, the best practice for these predictive tasks has been relatively unchanged and is still primarily based on variations of Gradient Boosted Decision Trees (GBDTs). Very recently, there has been renewed interest in developing state-of-the-art methods for tabular data based on recent developments in neural networks and feature learning methods. In this work, we introduce xRFM, an algorithm that combines feature learning kernel machines with a tree structure to both adapt to the local structure of the data and scale to essentially unlimited amounts of training data.
We show that compared to $31$ other methods, including recently introduced tabular foundation models (TabPFNv2) and GBDTs, xRFM achieves best performance across $100$ regression datasets and is competitive to the best methods across $200$ classification datasets outperforming GBDTs. Additionally, xRFM provides interpretability natively through the Average Gradient Outer Product.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 18:08:04 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Beaglehole",
"Daniel",
""
],
[
"Holzmüller",
"David",
""
],
[
"Radhakrishnan",
"Adityanarayanan",
""
],
[
"Belkin",
"Mikhail",
""
]
] |
2508.10054
|
Yaoqian Li
|
Yaoqian Li, Xikai Yang, Dunyuan Xu, Yang Yu, Litao Zhao, Xiaowei Hu, Jinpeng Li, Pheng-Ann Heng
|
SurgPub-Video: A Comprehensive Surgical Video Dataset for Enhanced Surgical Intelligence in Vision-Language Model
| null | null | null | null |
q-bio.OT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Vision-Language Models (VLMs) have shown significant potential in surgical scene analysis, yet existing models are limited by frame-level datasets and lack high-quality video data with procedural surgical knowledge. To address these challenges, we make the following contributions: (i) SurgPub-Video, a comprehensive dataset of over 3,000 surgical videos and 25 million annotated frames across 11 specialties, sourced from peer-reviewed clinical journals, (ii) SurgLLaVA-Video, a specialized VLM for surgical video understanding, built upon the TinyLLaVA-Video architecture that supports both video-level and frame-level inputs, and (iii) a video-level surgical Visual Question Answering (VQA) benchmark, covering diverse 11 surgical specialities, such as vascular, cardiology, and thoracic. Extensive experiments, conducted on the proposed benchmark and three additional surgical downstream tasks (action recognition, skill assessment, and triplet recognition), show that SurgLLaVA-Video significantly outperforms both general-purpose and surgical-specific VLMs with only three billion parameters. The dataset, model, and benchmark will be released to enable further advancements in surgical video understanding.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 18:23:13 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Li",
"Yaoqian",
""
],
[
"Yang",
"Xikai",
""
],
[
"Xu",
"Dunyuan",
""
],
[
"Yu",
"Yang",
""
],
[
"Zhao",
"Litao",
""
],
[
"Hu",
"Xiaowei",
""
],
[
"Li",
"Jinpeng",
""
],
[
"Heng",
"Pheng-Ann",
""
]
] |
2508.10055
|
Alokesh Manna
|
Alokesh Manna and Sujit K. Ghosh
|
Bayesian Models for Joint Selection of Features and Auto-Regressive Lags: Theory and Applications in Environmental and Financial Forecasting
| null | null | null | null |
stat.ME cs.LG math.ST stat.ML stat.TH
|
http://creativecommons.org/publicdomain/zero/1.0/
|
We develop a Bayesian framework for variable selection in linear regression with autocorrelated errors, accommodating lagged covariates and autoregressive structures. This setting occurs in time series applications where responses depend on contemporaneous or past explanatory variables and persistent stochastic shocks, including financial modeling, hydrological forecasting, and meteorological applications requiring temporal dependency capture. Our methodology uses hierarchical Bayesian models with spike-and-slab priors to simultaneously select relevant covariates and lagged error terms. We propose an efficient two-stage MCMC algorithm separating sampling of variable inclusion indicators and model parameters to address high-dimensional computational challenges. Theoretical analysis establishes posterior selection consistency under mild conditions, even when candidate predictors grow exponentially with sample size, common in modern time series with many potential lagged variables. Through simulations and real applications (groundwater depth prediction, S&P 500 log returns modeling), we demonstrate substantial gains in variable selection accuracy and predictive performance. Compared to existing methods, our framework achieves lower MSPE, improved true model component identification, and greater robustness with autocorrelated noise, underscoring practical utility for model interpretation and forecasting in autoregressive settings.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 18:44:36 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Manna",
"Alokesh",
""
],
[
"Ghosh",
"Sujit K.",
""
]
] |
2508.10056
|
Aske Nord Raahauge
|
Aske Nord Raahauge, Martin Bom Marchioro, Rasmus Ross Nylandsted
|
Approximating Entanglement Based on Abstract Interpretation
|
13 pages, no figures
| null | null | null |
quant-ph cs.ET
|
http://creativecommons.org/licenses/by/4.0/
|
Entanglement is a fundamental property of quantum systems, essential for non-trivial quantum programs. Identifying when qubits become entangled is critical for circuit optimization, and for arguing for the correctness of quantum algorithms. This paper presents a static analysis method for approximating entanglement by extending an already existing abstract interpretation, thus avoiding the exponential slowdown of an exact analysis. The approach is shown to be sound and an implementation is provided in Standard ML with linear-time scalability.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 19:15:40 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Raahauge",
"Aske Nord",
""
],
[
"Marchioro",
"Martin Bom",
""
],
[
"Nylandsted",
"Rasmus Ross",
""
]
] |
2508.10057
|
Christopher Pinier
|
Christopher Pinier, Sonia Acu\~na Vargas, Mariia Steeghs-Turchina, Dora Matzke, Claire E. Stevenson, Michael D. Nunez
|
Large Language Models Show Signs of Alignment with Human Neurocognition During Abstract Reasoning
|
Presented at the 8th Annual Conference on Cognitive Computational Neuroscience (August 12-15, 2025; Amsterdam, The Netherlands); 20 pages, 11 figures
| null | null | null |
q-bio.NC cs.AI cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
This study investigates whether large language models (LLMs) mirror human neurocognition during abstract reasoning. We compared the performance and neural representations of human participants with those of eight open-source LLMs on an abstract-pattern-completion task. We leveraged pattern type differences in task performance and in fixation-related potentials (FRPs) as recorded by electroencephalography (EEG) during the task. Our findings indicate that only the largest tested LLMs (~70 billion parameters) achieve human-comparable accuracy, with Qwen-2.5-72B and DeepSeek-R1-70B also showing similarities with the human pattern-specific difficulty profile. Critically, every LLM tested forms representations that distinctly cluster the abstract pattern categories within their intermediate layers, although the strength of this clustering scales with their performance on the task. Moderate positive correlations were observed between the representational geometries of task-optimal LLM layers and human frontal FRPs. These results consistently diverged from comparisons with other EEG measures (response-locked ERPs and resting EEG), suggesting a potential shared representational space for abstract patterns. This indicates that LLMs might mirror human brain mechanisms in abstract reasoning, offering preliminary evidence of shared principles between biological and artificial intelligence.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 21:38:46 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Pinier",
"Christopher",
""
],
[
"Vargas",
"Sonia Acuña",
""
],
[
"Steeghs-Turchina",
"Mariia",
""
],
[
"Matzke",
"Dora",
""
],
[
"Stevenson",
"Claire E.",
""
],
[
"Nunez",
"Michael D.",
""
]
] |
2508.10058
|
Kimberly Ayers
|
Kimberly Ayers, Maxwell Kooiker
|
A Structural Analysis of Population Graphs
| null | null | null | null |
q-bio.QM
|
http://creativecommons.org/licenses/by/4.0/
|
The format of graphing algorithms for genomic data has been a debate in recent biotechnology. In this paper, we discuss the construction of population graphs using said genomic data. We first examine the GENPOFAD distance measurement, developed by Joly et. al., and prove that this constitutes a metric function. We develop an algorithm to construct graphs to visualize the relationships between individuals in a population. We then provide a statistical analysis of these simulated population graphs, and show that they are distinct from randomly generated graphs, and also show differences from small-world graphs.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 21:51:08 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Ayers",
"Kimberly",
""
],
[
"Kooiker",
"Maxwell",
""
]
] |
2508.10059
|
Yueke Zhang
|
Yueke Zhang, Yifan Zhang, Kevin Leach, Yu Huang
|
FormalGrad: Integrating Formal Methods with Gradient-Based LLM Refinement
|
6 Pages
| null | null | null |
cs.SE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
While Large Language Models (LLMs) have demonstrated remarkable capabilities in code generation, they often produce solutions that lack guarantees of correctness, robustness, and efficiency. The limitation is acute in domains requiring strict constraints. FormalGrad introduces a principled framework that integrates formal methods directly into an iterative LLM-based generation loop. It uniquely treats code as a differentiable variable, converting structured feedback and formal constraints into a textual pseudo-gradient. This gradient guides the model to iteratively refine solutions, ensuring they are not only functional but also robust and formally justified. We evaluate FormalGrad on the HumanEval, HumanEval+, and LiveCodeBench benchmarks. Our implementation outperforms strong baselines, achieving an absolute improvement of up to 27% on HumanEval and a 41% relative improvement on the challenging LiveCodeBench V6. FormalGrad generates formally justified code that is robust and efficient, paving the way for reliable AI-assisted software development in high-stakes applications.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 22:03:54 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Zhang",
"Yueke",
""
],
[
"Zhang",
"Yifan",
""
],
[
"Leach",
"Kevin",
""
],
[
"Huang",
"Yu",
""
]
] |
2508.10060
|
Amy Armento Lee
|
Amy Armento Lee, Narayan Hegde, Nina Deliu, Emily Rosenzweig, Arun Suggala, Sriram Lakshminarasimhan, Qian He, John Hernandez, Martin Seneviratne, Rahul Singh, Pradnesh Kalkar, Karthikeyan Shanmugam, Aravindan Raghuveer, Abhimanyu Singh, My Nguyen, James Taylor, Jatin Alla, Sofia S. Villar, Hulya Emir-Farinas
|
A Personalized Exercise Assistant using Reinforcement Learning (PEARL): Results from a four-arm Randomized-controlled Trial
| null | null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Consistent physical inactivity poses a major global health challenge. Mobile health (mHealth) interventions, particularly Just-in-Time Adaptive Interventions (JITAIs), offer a promising avenue for scalable, personalized physical activity (PA) promotion. However, developing and evaluating such interventions at scale, while integrating robust behavioral science, presents methodological hurdles. The PEARL study was the first large-scale, four-arm randomized controlled trial to assess a reinforcement learning (RL) algorithm, informed by health behavior change theory, to personalize the content and timing of PA nudges via a Fitbit app.
We enrolled and randomized 13,463 Fitbit users into four study arms: control, random, fixed, and RL. The control arm received no nudges. The other three arms received nudges from a bank of 155 nudges based on behavioral science principles. The random arm received nudges selected at random. The fixed arm received nudges based on a pre-set logic from survey responses about PA barriers. The RL group received nudges selected by an adaptive RL algorithm. We included 7,711 participants in primary analyses (mean age 42.1, 86.3% female, baseline steps 5,618.2).
We observed an increase in PA for the RL group compared to all other groups from baseline to 1 and 2 months. The RL group had significantly increased average daily step count at 1 month compared to all other groups: control (+296 steps, p=0.0002), random (+218 steps, p=0.005), and fixed (+238 steps, p=0.002). At 2 months, the RL group sustained a significant increase compared to the control group (+210 steps, p=0.0122). Generalized estimating equation models also revealed a sustained increase in daily steps in the RL group vs. control (+208 steps, p=0.002). These findings demonstrate the potential of a scalable, behaviorally-informed RL approach to personalize digital health interventions for PA.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 22:44:33 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Lee",
"Amy Armento",
""
],
[
"Hegde",
"Narayan",
""
],
[
"Deliu",
"Nina",
""
],
[
"Rosenzweig",
"Emily",
""
],
[
"Suggala",
"Arun",
""
],
[
"Lakshminarasimhan",
"Sriram",
""
],
[
"He",
"Qian",
""
],
[
"Hernandez",
"John",
""
],
[
"Seneviratne",
"Martin",
""
],
[
"Singh",
"Rahul",
""
],
[
"Kalkar",
"Pradnesh",
""
],
[
"Shanmugam",
"Karthikeyan",
""
],
[
"Raghuveer",
"Aravindan",
""
],
[
"Singh",
"Abhimanyu",
""
],
[
"Nguyen",
"My",
""
],
[
"Taylor",
"James",
""
],
[
"Alla",
"Jatin",
""
],
[
"Villar",
"Sofia S.",
""
],
[
"Emir-Farinas",
"Hulya",
""
]
] |
2508.10061
|
Fu Wanjia
|
Wanjia Fu and Yingying Ma and Hanzhong Liu
|
Regression adjustment in covariate-adaptive randomized experiments with missing covariates
| null | null | null | null |
stat.ME math.ST stat.TH
|
http://creativecommons.org/licenses/by/4.0/
|
Covariate-adaptive randomization is widely used in clinical trials to balance prognostic factors, and regression adjustments are often adopted to further enhance the estimation and inference efficiency. In practice, the covariates may contain missing values. Various methods have been proposed to handle the covariate missing problem under simple randomization. However, the statistical properties of the resulting average treatment effect estimators under stratified randomization, or more generally, covariate-adaptive randomization, remain unclear. To address this issue, we investigate the asymptotic properties of several average treatment effect estimators obtained by combining commonly used missingness processing procedures and regression adjustment methods. Moreover, we derive consistent variance estimators to enable valid inferences. Finally, we conduct a numerical study to evaluate the finite-sample performance of the considered estimators under various sample sizes and numbers of covariates and provide recommendations accordingly. Our analysis is model-free, meaning that the conclusions remain asymptotically valid even in cases of misspecification of the regression model.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 03:07:29 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Fu",
"Wanjia",
""
],
[
"Ma",
"Yingying",
""
],
[
"Liu",
"Hanzhong",
""
]
] |
2508.10062
|
Ying Dai
|
Ying Dai, Na Ouyang
|
Excessive Screen Time is Associated with Mental Health Problems and ADHD in US Children and Adolescents: Physical Activity and Sleep as Parallel Mediators
| null | null | null | null |
q-bio.NC
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
To examine associations between screen time and anxiety, depression, behavior or conduct problems, and ADHD among children and adolescents during the pandemic, and to assess whether physical activity, sleep duration, and bedtime regularity mediate these associations. Data from 50231 US children and adolescents aged 6 to 17 years in the 2020 to 2021 National Survey of Childrens Health were analyzed. Exact natural effect models and structural equation modeling assessed mediation by physical activity, short sleep duration, and irregular bedtime. We found that daily screen time equal or more than 4 hours was linked to higher risks of anxiety (aOR = 1.45, 95% CI 1.32, 1.58), depression (aOR = 1.65, 95% CI 1.41, 1.93), behavior or conduct problems (aOR = 1.17, 95% CI 1.05, 1.30), and ADHD (aOR = 1.21, 95% CI 1.11, 1.33). Physical activity accounted for 30.2% to 39.3% of the association, irregular bedtime for 18.2% to 25.7%, and short sleep duration for 2.77% to 7.34%. Excessive screen time was associated with poorer mental health and ADHD, partly explained by reduced physical activity, irregular bedtime, and insufficient sleep. Interventions should promote physical activity, regular sleep routines, and adequate sleep duration to effectively mitigate mental health issues and ADHD among children and adolescents.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 03:13:44 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Dai",
"Ying",
""
],
[
"Ouyang",
"Na",
""
]
] |
2508.10063
|
Yuntian Xia
|
Steven Klee and Yuntian Xia
|
Measuring Time Series Forecast Stability for Demand Planning
|
6 pages, 3 figures; KDD '25
| null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Time series forecasting is a critical first step in generating demand plans for supply chains. Experiments on time series models typically focus on demonstrating improvements in forecast accuracy over existing/baseline solutions, quantified according to some accuracy metric. There is no doubt that forecast accuracy is important; however in production systems, demand planners often value consistency and stability over incremental accuracy improvements. Assuming that the inputs have not changed significantly, forecasts that vary drastically from one planning cycle to the next require high amounts of human intervention, which frustrates demand planners and can even cause them to lose trust in ML forecasting models. We study model-induced stochasticity, which quantifies the variance of a set of forecasts produced by a single model when the set of inputs is fixed. Models with lower variance are more stable.
Recently the forecasting community has seen significant advances in forecast accuracy through the development of deep machine learning models for time series forecasting. We perform a case study measuring the stability and accuracy of state-of-the-art forecasting models (Chronos, DeepAR, PatchTST, Temporal Fusion Transformer, TiDE, and the AutoGluon best quality ensemble) on public data sets from the M5 competition and Favorita grocery sales. We show that ensemble models improve stability without significantly deteriorating (or even improving) forecast accuracy. While these results may not be surprising, the main point of this paper is to propose the need for further study of forecast stability for models that are being deployed in production systems.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 04:21:37 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Klee",
"Steven",
""
],
[
"Xia",
"Yuntian",
""
]
] |
2508.10064
|
Xia Chen
|
Xia Chen
|
Dynamical Alignment: A Principle for Adaptive Neural Computation
|
16 pages, 10 figures;
| null | null | null |
q-bio.NC cs.LG
|
http://creativecommons.org/licenses/by-sa/4.0/
|
The computational capabilities of a neural network are widely assumed to be determined by its static architecture. Here we challenge this view by establishing that a fixed neural structure can operate in fundamentally different computational modes, driven not by its structure but by the temporal dynamics of its input signals. We term this principle 'Dynamical Alignment'.
Applying this principle offers a novel resolution to the long-standing paradox of why brain-inspired spiking neural networks (SNNs) underperform. By encoding static input into controllable dynamical trajectories, we uncover a bimodal optimization landscape with a critical phase transition governed by phase space volume dynamics. A 'dissipative' mode, driven by contracting dynamics, achieves superior energy efficiency through sparse temporal codes. In contrast, an 'expansive' mode, driven by expanding dynamics, unlocks the representational power required for SNNs to match or even exceed their artificial neural network counterparts on diverse tasks, including classification, reinforcement learning, and cognitive integration.
We find this computational advantage emerges from a timescale alignment between input dynamics and neuronal integration. This principle, in turn, offers a unified, computable perspective on long-observed dualities in neuroscience, from stability-plasticity dilemma to segregation-integration dynamic. It demonstrates that computation in both biological and artificial systems can be dynamically sculpted by 'software' on fixed 'hardware', pointing toward a potential paradigm shift for AI research: away from designing complex static architectures and toward mastering adaptive, dynamic computation principles.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 06:35:57 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Chen",
"Xia",
""
]
] |
2508.10065
|
Yuhao Sun
|
Yuhao Sun, Yihua Zhang, Gaowen Liu, Hongtao Xie, Sijia Liu
|
Invisible Watermarks, Visible Gains: Steering Machine Unlearning with Bi-Level Watermarking Design
|
Accepted by ICCV 2025
| null | null | null |
cs.CR cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
With the increasing demand for the right to be forgotten, machine unlearning (MU) has emerged as a vital tool for enhancing trust and regulatory compliance by enabling the removal of sensitive data influences from machine learning (ML) models. However, most MU algorithms primarily rely on in-training methods to adjust model weights, with limited exploration of the benefits that data-level adjustments could bring to the unlearning process. To address this gap, we propose a novel approach that leverages digital watermarking to facilitate MU by strategically modifying data content. By integrating watermarking, we establish a controlled unlearning mechanism that enables precise removal of specified data while maintaining model utility for unrelated tasks. We first examine the impact of watermarked data on MU, finding that MU effectively generalizes to watermarked data. Building on this, we introduce an unlearning-friendly watermarking framework, termed Water4MU, to enhance unlearning effectiveness. The core of Water4MU is a bi-level optimization (BLO) framework: at the upper level, the watermarking network is optimized to minimize unlearning difficulty, while at the lower level, the model itself is trained independently of watermarking. Experimental results demonstrate that Water4MU is effective in MU across both image classification and image generation tasks. Notably, it outperforms existing methods in challenging MU scenarios, known as "challenging forgets".
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 08:05:47 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Sun",
"Yuhao",
""
],
[
"Zhang",
"Yihua",
""
],
[
"Liu",
"Gaowen",
""
],
[
"Xie",
"Hongtao",
""
],
[
"Liu",
"Sijia",
""
]
] |
2508.10066
|
Javier R\'odenas
|
Javier Rodenas, Eduardo Aguilar, Petia Radeva
|
Stochastic-based Patch Filtering for Few-Shot Learning
|
CVPR Workshop MetaFood 2025
|
J. Rodenas Cumplido, E. Aguilar, and P. Radeva, "Slot Attention-based Feature Filtering for Few-Shot Learning," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR) Workshops, 2025, pp. 440-449
| null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Food images present unique challenges for few-shot learning models due to their visual complexity and variability. For instance, a pasta dish might appear with various garnishes on different plates and in diverse lighting conditions and camera perspectives. This problem leads to losing focus on the most important elements when comparing the query with support images, resulting in misclassification. To address this issue, we propose Stochastic-based Patch Filtering for Few-Shot Learning (SPFF) to attend to the patch embeddings that show greater correlation with the class representation. The key concept of SPFF involves the stochastic filtering of patch embeddings, where patches less similar to the class-aware embedding are more likely to be discarded. With patch embedding filtered according to the probability of appearance, we use a similarity matrix that quantifies the relationship between the query image and its respective support images. Through a qualitative analysis, we demonstrate that SPFF effectively focuses on patches where class-specific food features are most prominent while successfully filtering out non-relevant patches. We validate our approach through extensive experiments on few-shot classification benchmarks: Food-101, VireoFood-172 and UECFood-256, outperforming the existing SoA methods.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 11:05:05 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Rodenas",
"Javier",
""
],
[
"Aguilar",
"Eduardo",
""
],
[
"Radeva",
"Petia",
""
]
] |
2508.10067
|
Yoonhu Kim
|
Yoonhu Kim
|
Undecidability of Tiling the Plane with a Set of 5 Polyominoes
| null | null | null | null |
math.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we give a proof that it is undecidable whether a set of five polyominoes can tile the plane by translation. The proof involves a new method of labeling the edges of polyominoes, making it possible to assign whether two edges can match for any set of two edges chosen. This is achieved by dedicating 1 polyomino to the labeling process.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 11:24:47 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Kim",
"Yoonhu",
""
]
] |
2508.10068
|
Xiaohan Chen
|
Xiaohan Chen, Zhongying Pan, Quan Feng, Yu Tian, Shuqun Yang, Mengru Wang, Lina Gong, Yuxia Geng, Piji Li, Xiang Chen
|
SaraCoder: Orchestrating Semantic and Structural Cues for Profit-Oriented Repository-Level Code Completion
| null | null | null | null |
cs.SE cs.CL cs.IR cs.PL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Retrieval-augmented generation (RAG) for repository-level code completion commonly relies on superficial text similarity, leading to results plagued by semantic misguidance, redundancy, and homogeneity, while also failing to resolve external symbol ambiguity. To address these challenges, we introduce Saracoder, a Hierarchical Feature-Optimized retrieval framework. Its core Hierarchical Feature Optimization module systematically refines candidates by distilling deep semantic relationships, pruning exact duplicates, assessing structural similarity with a novel graph-based metric that weighs edits by their topological importance, and reranking results to maximize both relevance and diversity. Furthermore, an External-Aware Identifier Disambiguator module accurately resolves cross-file symbol ambiguity via dependency analysis. Extensive experiments on the challenging CrossCodeEval and RepoEval-Updated benchmarks demonstrate that Saracoder significantly outperforms existing baselines across multiple programming languages and models. Our work proves that systematically refining retrieval results across multiple dimensions provides a new paradigm for building more accurate and robust repository-level code completion systems.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 11:56:05 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Chen",
"Xiaohan",
""
],
[
"Pan",
"Zhongying",
""
],
[
"Feng",
"Quan",
""
],
[
"Tian",
"Yu",
""
],
[
"Yang",
"Shuqun",
""
],
[
"Wang",
"Mengru",
""
],
[
"Gong",
"Lina",
""
],
[
"Geng",
"Yuxia",
""
],
[
"Li",
"Piji",
""
],
[
"Chen",
"Xiang",
""
]
] |
2508.10069
|
Faizuddin Ahmed
|
Faizuddin Ahmed and Abdelmalek Bouzenada
|
Quantum-Corrected Deformation of RN-AdS Black Holes with a Cloud of Strings Immersed in Quintessential-like Fluid
|
16 pages; 9 figures; 4 tables
| null | null | null |
gr-qc hep-th
|
http://creativecommons.org/licenses/by/4.0/
|
Inspired by Loop Quantum Gravity (LQG), we investigate the Reissner-Nordstr\"om (RN) black hole (BH) solution coupled with a cloud of strings in an anti-de Sitter (AdS) background, surrounded by a quintessence-like fluid. We begin by analyzing the optical properties of the BH through null geodesic motion, deriving an effective potential that governs photon dynamics. This effective potential is central to understanding key phenomena such as photon trajectories, circular null orbits, photon sphere, and the resulting BH shadow. Subsequently, we study the dynamics of neutral test particles by deriving effective potential that describes their motion. Using the potential, we compute the specific energy and specific angular momentum of neutral particles in circular orbits around the BH and analyze the outcomes. We also investigate the innermost stable circular orbits (ISCO) and demonstrate how various geometrical and physical properties influence the radius of ISCO. Furthermore, we explore the thermodynamic properties of the BH solution by deriving key quantities such as the Hawking temperature, entropy, Gibbs free energy, internal energy, and specific heat capacity. Throughout the study, we demonstrate that the geodesic structure, scalar field behavior, and thermodynamic properties are significantly influenced by parameters such as the string cloud, quantum correction, electric charge, the surrounding quintessence-like fluid, and the AdS curvature radius.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 12:19:08 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Ahmed",
"Faizuddin",
""
],
[
"Bouzenada",
"Abdelmalek",
""
]
] |
2508.10070
|
Weitao Ye
|
Weitao Ye, Yungui Gong, Tomohiro Harada, Zhaofeng Kang, Kazunori Kohri, Daiki Saito and Chul-Moon Yoo
|
Primordial Black Hole Formation and Spin in Matter Domination Revisited
|
22 pages, 3 figures, 5 tables
| null | null |
NU-QG-9, RUP-25-19, KUNS-3070
|
gr-qc astro-ph.CO hep-ph hep-th
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this article, we calculate the mass distribution of primordial black holes (PBHs) formed in the matter-dominated (MD) era by the peak theory. We apply the Zel'dovich approximation to track the nonlinear evolution of overdensities and compute the PBH abundance and mass function by incorporating a PBH formation criterion based on the hoop conjecture. We find that the PBH abundance $\beta$ follows the scaling law $\beta \simeq A_\gamma \sigma_h^{*5}$ for $\sigma_h^*\ll 1$. Here, $\sigma_h^*$ is the quantity that characterizes the variance of the density fluctuation at the horizon entry. We also find that, in contrast to the previous estimates, the PBH spin is very small for $\sigma_h^*\ll 1$ but could be larger for larger $\sigma_h^*$ and broader power spectra. Finally, specializing to a monochromatic power spectrum, we prove analytically that the PBH mass distribution becomes effectively monochromatic and reveal that the resultant PBH abundance is approximately 19 times the previous prediction.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 13:03:42 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Ye",
"Weitao",
""
],
[
"Gong",
"Yungui",
""
],
[
"Harada",
"Tomohiro",
""
],
[
"Kang",
"Zhaofeng",
""
],
[
"Kohri",
"Kazunori",
""
],
[
"Saito",
"Daiki",
""
],
[
"Yoo",
"Chul-Moon",
""
]
] |
2508.10071
|
Jay Cunningham
|
Jay L. Cunningham, Kevin Zhongyang Shao, Rock Yuren Pang, Nathaniel Mengist
|
Advancing Data Equity: Practitioner Responsibility and Accountability in NLP Data Practices
|
10 pages, 6 Pages (References and Appendices). The archival version has been accepted to AAAI (AIES 2025) without the extended Appendices. This extended version includes Appendices
| null | null | null |
cs.CY cs.AI cs.HC
|
http://creativecommons.org/licenses/by/4.0/
|
While research has focused on surfacing and auditing algorithmic bias to ensure equitable AI development, less is known about how NLP practitioners - those directly involved in dataset development, annotation, and deployment - perceive and navigate issues of NLP data equity. This study is among the first to center practitioners' perspectives, linking their experiences to a multi-scalar AI governance framework and advancing participatory recommendations that bridge technical, policy, and community domains. Drawing on a 2024 questionnaire and focus group, we examine how U.S.-based NLP data practitioners conceptualize fairness, contend with organizational and systemic constraints, and engage emerging governance efforts such as the U.S. AI Bill of Rights. Findings reveal persistent tensions between commercial objectives and equity commitments, alongside calls for more participatory and accountable data workflows. We critically engage debates on data diversity and diversity washing, arguing that improving NLP equity requires structural governance reforms that support practitioner agency and community consent.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 13:14:43 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Cunningham",
"Jay L.",
""
],
[
"Shao",
"Kevin Zhongyang",
""
],
[
"Pang",
"Rock Yuren",
""
],
[
"Mengist",
"Nathaniel",
""
]
] |
2508.10072
|
Vladimir A. Petrov
|
Vladimir A. Petrov
|
Can the Odderon be "Supercritical"?
| null | null | null | null |
hep-ph
|
http://creativecommons.org/publicdomain/zero/1.0/
|
Based on the postulate that a $C$-odd exchange leads to repulsion in the particle-particle system (say, $pp$) and to attraction in the antiparticle-particle system (say, $\bar{p}p$), we show that the Odderon intercept cannot be greater than unity.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 14:16:22 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Petrov",
"Vladimir A.",
""
]
] |
2508.10073
|
Nuraly Adilkhan
|
Nuraly Adilkhan
|
Fragility Spectrum: Measuring Resilience in Model-Theoretic Properties under Language Expansions
| null | null | null | null |
math.LO
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
We introduce the fragility spectrum, a quantitative framework to measure the resilience of model-theoretic properties (e.g., stability, NIP, NTP$_2$, decidability) under language expansions. The core is the fragility index $\operatorname{frag}(T, P \to Q)$, quantifying the minimal expansion needed to degrade from property $P$ to $Q$. We axiomatize fragility operators, prove stratification theorems, identify computational, geometric, and combinatorial collapse modes, and position it within Shelah's hierarchy. Examples include ACF$_0$ (infinite fragility for stability) and $\operatorname{Th}(\mathbb{Q}, +)$ (fragility 1 for $\omega$-stability). Connections to DOP, ranks, and external definability refine classifications. Extended proofs, applications to other logics, and open problems enhance the discourse.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 15:50:05 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Adilkhan",
"Nuraly",
""
]
] |
2508.10074
|
Ruofan Lu
|
Ruofan Lu, Yintong Huo, Meng Zhang, Yichen Li, Michael R. Lyu
|
Next Edit Prediction: Learning to Predict Code Edits from Context and Interaction History
| null | null | null | null |
cs.SE cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The rapid advancement of large language models (LLMs) has led to the widespread adoption of AI-powered coding assistants integrated into a development environment. On one hand, low-latency code completion offers completion suggestions but is fundamentally constrained to the cursor's current position. On the other hand, chat-based editing can perform complex modifications, yet forces developers to stop their work, describe the intent in natural language, which causes a context-switch away from the code. This creates a suboptimal user experience, as neither paradigm proactively predicts the developer's next edit in a sequence of related edits. To bridge this gap and provide the seamless code edit suggestion, we introduce the task of Next Edit Prediction, a novel task designed to infer developer intent from recent interaction history to predict both the location and content of the subsequent edit. Specifically, we curate a high-quality supervised fine-tuning dataset and an evaluation benchmark for the Next Edit Prediction task. Then, we conduct supervised fine-tuning on a series of models and performed a comprehensive evaluation of both the fine-tuned models and other baseline models, yielding several novel findings. This work lays the foundation for a new interaction paradigm that proactively collaborate with developers by anticipating their following action, rather than merely reacting to explicit instructions.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 15:52:03 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Lu",
"Ruofan",
""
],
[
"Huo",
"Yintong",
""
],
[
"Zhang",
"Meng",
""
],
[
"Li",
"Yichen",
""
],
[
"Lyu",
"Michael R.",
""
]
] |
2508.10075
|
Nesar Ramachandra
|
Nesar Ramachandra, Yuan-Sen Ting, Zechang Sun, Azton Wells, Salman Habib
|
Teaching LLMs to Speak Spectroscopy
|
6 pages, 1 figure, Accepted to the Machine Learning for Astrophysics (ML4Astro) Colocated Workshop at ICML 2025
| null | null | null |
astro-ph.IM
|
http://creativecommons.org/licenses/by/4.0/
|
Pre-trained Large Language Models (LLMs) have revolutionized text processing, yet adapting Transformer-based neural networks to non-textual scientific modalities typically requires specialized architectures and extensive computational resources. We demonstrate that LLaMA-3.1-8B can be efficiently repurposed to predict galaxy redshifts from spectroscopic data through Low-Rank Adaptation (LoRA), achieving competitive performance while preserving its linguistic capabilities. Using only 16 GPU-hours and adapting 0.04% of model parameters, our approach achieves a mean absolute error of 0.04 in redshift prediction while retaining over 85% of performance on AstroBench and 89% on general QA tasks from eval-harness. This minimal-effort adaptation--requiring only simple standard fine-tuning APIs--lowers barriers to entry for domain scientists and enables integrated agentic workflows where a single model handles both spectroscopic data for quantitative analysis and natural language for reasoning.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 16:27:12 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Ramachandra",
"Nesar",
""
],
[
"Ting",
"Yuan-Sen",
""
],
[
"Sun",
"Zechang",
""
],
[
"Wells",
"Azton",
""
],
[
"Habib",
"Salman",
""
]
] |
2508.10076
|
Lukas Devos
|
Lukas Devos and Jutho Haegeman
|
TensorKit.jl: A Julia package for large-scale tensor computations, with a hint of category theory
|
69 pages, 4 figures
| null | null | null |
cs.MS cond-mat.str-el quant-ph
|
http://creativecommons.org/licenses/by-sa/4.0/
|
TensorKit.jl is a Julia-based software package for tensor computations, especially focusing on tensors with internal symmetries. This paper introduces the design philosophy, core functionalities, and distinctive features, including how to handle abelian, non-abelian, and anyonic symmetries through the ``TensorMap'' type. We highlight the software's flexibility, performance, and its capability to extend to new tensor types and symmetries, illustrating its practical applications through select case studies.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 16:35:11 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Devos",
"Lukas",
""
],
[
"Haegeman",
"Jutho",
""
]
] |
2508.10077
|
Sufiyan Mallu
|
Peter Dankelmann, Sonwabile Mafunda, Sufiyan Mallu
|
Proximity and Radius in Outerplanar Graphs with Bounded Faces
| null | null | null | null |
math.CO
|
http://creativecommons.org/licenses/by/4.0/
|
Let $G$ be a finite, connected graph and $v$ a vertex of $G$. The average distance and the eccentricity of $v$ in $G$ are defined as the arithmetic mean and the maximum, respectively, of the distances from $v$ to all other vertices of $G$. The proximity of $G$ and the radius of $G$ are defined as the minimum of the average distances and the eccentricities over all vertices of $G$.
In this paper, we establish an upper bound on the proximity of a $2$-connected outerplanar graphs in terms of order and maximum face length. This bound is sharp apart from a small additive constant.
It is known that the radius of a maximal outerplanar graph is at most $\lfloor \frac{n}{4} \rfloor +1$. In the second part of this paper we show that this bound on the radius holds for a much larger subclass of outerplanar graphs, for all $2$-connected outerplanar graphs of order $n$ whose maximum face length does not exceed $\frac{n+2}{4}$.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 16:37:15 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Dankelmann",
"Peter",
""
],
[
"Mafunda",
"Sonwabile",
""
],
[
"Mallu",
"Sufiyan",
""
]
] |
2508.10078
|
Sufiyan Mallu
|
Peter Dankelmann, Sonwabile Mafunda, Sufiyan Mallu
|
On Proximity and other Distance Parameters in Planar Graphs
| null | null | null | null |
math.CO
|
http://creativecommons.org/licenses/by/4.0/
|
Let $G$ be a connected graph. The average distance of a vertex $v$ of $G$ is the arithmetic mean of the distances from $v$ to all other vertices of $G$. The proximity and remoteness of $G$ are defined as the minimum and maximum, respectively, of the average distances of the vertices of $G$.
It was shown by Aouchiche and Hansen [Proximity and remoteness in graphs: bounds and conjectures, Networks 58 no.\ 2 (2011)] that for a connected graph of order $n$, the difference between remoteness and proximity and the difference between radius and proximity are bounded from above by about $\frac{n}{4}$, and the difference between diameter and proximity is bounded from above by about $\frac{3}{4}n$.
In this paper, we show that all three bounds can be improved significantly for maximal planar graphs, and for graphs of given connectivity.
We show that in maximal planar graphs the above bound on the difference between radius and proximity can be improved to about $\frac{1}{12}n$, and further to about $\frac{1}{16}n$ and $\frac{1}{20}n$ if the graphs is, in addition, $4$-connected or $5$-connected, respectively. Similar improvements are shown for quadrangulations, and for maximal outerplanar graphs. We further show that the above bound on the difference between remoteness and proximity can be improved to about $\frac{1}{4\kappa}n$ if $G$ is $\kappa$-connected. Finally, we improve the bound on the difference between diameter and proximity to about $\frac{3}{4\kappa}n$ if $G$ is $\kappa$-connected. We present graphs that demonstrate that our bounds are either sharp, or sharp apart from an additive constant, even if restricted to planar graphs.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 16:50:46 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Dankelmann",
"Peter",
""
],
[
"Mafunda",
"Sonwabile",
""
],
[
"Mallu",
"Sufiyan",
""
]
] |
2508.10079
|
Andrada Pojar
|
Andrada Pojar
|
Nonderogatory matrices as sums of $p$-potent and nilpotent matrices
| null | null | null | null |
math.RA
|
http://creativecommons.org/licenses/by/4.0/
|
We prove that every nonderogatory matrix $A$ over a field of positive odd characteristic $p,$ that is sum of a $p$-potent matrix and a nilpotent matrix, has a decomposition $A=E+V,$ such that $E^p=E,$ and $V^3=0.$
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 17:11:58 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Pojar",
"Andrada",
""
]
] |
2508.10080
|
Jiaqiang Yan
|
Jiaqiang Yan, David Singh, Bayram Saparov, Huibo Cao, Yejun Feng, Jinguang Cheng, Yoshia Uwatoko, and David Mandrus
|
Structural transition and possible pressure-induced superconductivity in a suboxide La$_5$Pb$_3$O
|
9 figures, 9 pages
| null | null | null |
cond-mat.supr-con cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by/4.0/
|
Here we report a structural phase transition and its possible competition with superconductivity in the suboxide La$_5$Pb$_3$O. Upon cooling through $T_t$ = 225 K, La$_5$Pb$_3$O transforms from a high-temperature I4/mcm to a low-temperature P4/ncc structure in which La - Pb dimerization along the c-axis occurs. This transition is accompanied by anomalies in the temperature dependence of electrical resistivity and specific heat. High-pressure electrical transport measurements reveal that hydrostatic pressure suppresses the structural transition and possibly induces superconductivity with a maximum superconducting temperature of 10 K. Density functional theory calculations show minimal changes in the electronic density of states and no gap opening at $E_F$ across $T_t$, suggesting that the transition is driven by bonding effects rather than Fermi surface instability. These findings establish La$_5$Pb$_3$O as a promising platform for exploring the interplay between weak structural transitions and superconductivity.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 17:29:46 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Yan",
"Jiaqiang",
""
],
[
"Singh",
"David",
""
],
[
"Saparov",
"Bayram",
""
],
[
"Cao",
"Huibo",
""
],
[
"Feng",
"Yejun",
""
],
[
"Cheng",
"Jinguang",
""
],
[
"Uwatoko",
"Yoshia",
""
],
[
"Mandrus",
"David",
""
]
] |
2508.10081
|
Andrada Pojar
|
Andrada Pojar
|
$m$-nil-clean nonderogatory matrices
| null | null | null | null |
math.RA
|
http://creativecommons.org/licenses/by/4.0/
|
We prove that if $\mathbb{F}$ is a field of positive odd characteristic $p,$ and $m,$ and $n$ are positive integers such that $m\geq2,$ and $n\leq p,$ every $n\times n$ nonderogatory matrix $A\in \mathbb{M}_n(\mathbb{F})$ which is sum of $m$ idempotents and a nilpotent, has a decomposition $A=E_1+E_2+\dots+E_m+V,$ such that $E_i^2=E_i,$ for every $i\in \{1,\dots,m\},$ and $V^{[\frac{p-2}{m}]+2}=0.$
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 17:32:37 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Pojar",
"Andrada",
""
]
] |
2508.10082
|
Bin Hu
|
Yousra Hassan Alsaad Almeshale, Abdulelah Hassan Almeshali, Omar Alsaddique, Noura Jandali, Nadeen Garaween, and Bin Hu
|
Developing an Inhaled NEU1 Inhibitor for Cystic Fibrosis via Pharmacokinetic and Biophysical Modeling
|
4 figures, 5 tables
| null | null | null |
q-bio.QM
|
http://creativecommons.org/licenses/by/4.0/
|
Background: Cystic fibrosis (CF) airway mucus exhibits reduced mucin sialylation, increasing viscosity and impairing mucociliary clearance (MCC). NEU1 inhibition has been proposed to restore MCC, but its quantitative pharmacokinetic and rheological effects, particularly with inhaled delivery, remain uncharacterized.
Objective: To develop an integrated pharmacokinetic/pharmacodynamic (PK/PD) and biophysical model to assess the efficacy of an inhaled NEU1 inhibitor.
Methods: Empirical and preclinical NEU1 inhibition data were combined with inhalation PK/PD modeling and a biophysical viscosity framework linking mucin sialylation and extracellular DNA. Synthetic cohort simulations (N = 200) were reconciled with empirical PK benchmarks using Latin hypercube parameter sampling. Cross-validation, hold-out testing, and causal inference methods (inverse probability of treatment weighting and targeted maximum likelihood estimation) quantified predicted effects on lung function (delta FEV1).
Results: With reconciled parameters (F_dep = 0.12; k_abs = 0.21 per hour; k_muc = 0.24 per hour), epithelial lining fluid drug levels reached a peak concentration of 7.5 micromolar (95 percent CI: 6 to 10 micromolar), achieving IC50 coverage for approximately 10 hours per day and greater than 80 percent modeled NEU1 inhibition. Predicted mucus viscosity reduction averaged 25 to 28 percent. Causal inference estimated delta FEV1 improvement of +0.13 liters (95 percent CI: 0.10 to 0.15 liters), with about 70 percent mediated via MCC.
Conclusions: Empirically anchored PK/PD and biophysical modeling support the feasibility of inhaled NEU1 inhibition as a rheology-targeting strategy in CF, projecting clinically realistic efficacy while maintaining pharmacological viability. This calibrated proof of concept warrants in vivo validation in CF models.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 17:42:14 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Almeshale",
"Yousra Hassan Alsaad",
""
],
[
"Almeshali",
"Abdulelah Hassan",
""
],
[
"Alsaddique",
"Omar",
""
],
[
"Jandali",
"Noura",
""
],
[
"Garaween",
"Nadeen",
""
],
[
"Hu",
"Bin",
""
]
] |
2508.10083
|
Art Owen
|
Art B. Owen
|
Better bootstrap t confidence intervals for the mean
| null | null | null | null |
math.ST stat.CO stat.TH
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This article explores combinations of weighted bootstraps, like the Bayesian bootstrap, with the bootstrap $t$ method for setting approximate confidence intervals for the mean of a random variable in small samples. For this problem the usual bootstrap $t$ has good coverage but provides intervals with long and highly variable lengths. Those intervals can have infinite length not just for tiny $n$, when the data have a discrete distribution. The BC$_a$ bootstrap produces shorter intervals but tends to severely under-cover the mean. Bootstrapping the studentized mean with weights from a Beta$(1/2,3/2)$ distribution is shown to attain second order accuracy. It never yields infinite length intervals and the mean square bootstrap $t$ statistic is finite when there are at least three distinct values in the data, or two distinct values appearing at least three times each. In a range of small sample settings, the beta bootstrap $t$ intervals have closer to nominal coverage than the BC$_a$ and shorter length than the multinomial ootstrap $t$. The paper includes a lengthy discussion of the difficulties in constructing a utility function to evaluate nonparametric approximate confidence intervals.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 17:55:00 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Owen",
"Art B.",
""
]
] |
2508.10084
|
Jumpei Tanaka
|
Jumpei Tanaka
|
Types of graded tensor products of graded von Neumann algebras
|
24 pages
| null | null | null |
math.OA
|
http://creativecommons.org/licenses/by/4.0/
|
There is a famous multiplication table of types of tensor product of two von Neumann algebras. We filled out the multiplication table of graded tensor product of two graded von Neumann algebras in special cases.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 17:57:59 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Tanaka",
"Jumpei",
""
]
] |
2508.10085
|
Oliver Payne
|
Andrea Fontanella and Oliver Payne
|
A Carroll Limit of AdS/CFT: A Triality with Flat Space Holography?
|
5p + supplemental material, 1 fig
| null | null | null |
hep-th
|
http://creativecommons.org/licenses/by/4.0/
|
We construct a novel holographic duality by taking a Carrollian limit of the AdS/CFT correspondence, relating string theory in a Carroll bulk geometry to a Carroll $\mathcal{N}=4$ Super Yang-Mills theory. We further propose the existence of an underlying triality connecting Carroll string theory, relativistic string theory in flat spacetime, and Carroll gauge theory. Finally, we analyse the symmetries of the Carroll gauge theory, showing that they are non-linearly realised and do not close under ordinary Lie brackets.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:00 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Fontanella",
"Andrea",
""
],
[
"Payne",
"Oliver",
""
]
] |
2508.10086
|
Daniil Rabinovich
|
Daniil Rabinovich, Andrey Kardashin, and Soumik Adhikary
|
On the role of overparametrization in Quantum Approximate Optimization
|
11 pages, 5 figures
| null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Variational quantum algorithms have emerged as a cornerstone of contemporary quantum algorithms research. While they have demonstrated considerable promise in solving problems of practical interest, efficiently determining the minimal quantum resources necessary to obtain such a solution remains an open question. In this work, inspired by concepts from classical machine learning, we investigate the impact of overparameterization on the performance of variational algorithms. Our study focuses on the quantum approximate optimization algorithm (QAOA) -- a prominent variational quantum algorithm designed to solve combinatorial optimization problems. We investigate if circuit overparametrization is necessary and sufficient to solve such problems in QAOA, considering two representative problems -- MAX-CUT and MAX-2-SAT. For MAX-CUT we observe that overparametriation is both sufficient and (statistically) necessary for attaining exact solutions, as confirmed numerically for up to $20$ qubits. In fact, for MAX-CUT on 2-regular graphs we show the necessity to be exact, based on the analytically found optimal depth. In sharp contrast, for MAX-2-SAT, underparametrized circuits suffice to solve most instances. This result highlights the potential of QAOA in the underparametrized regime, supporting its utility for current noisy devices.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:00 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Rabinovich",
"Daniil",
""
],
[
"Kardashin",
"Andrey",
""
],
[
"Adhikary",
"Soumik",
""
]
] |
2508.10087
|
Davide Passaro
|
Shimal Harichurn, Mrunmay Jagadale, Dmitry Noshchenko, Davide Passaro
|
$c_\text{eff}$ from Surgery and Modularity
|
50 pages, 7 tables, 10 figures
| null | null |
DIAS-STP-25-20
|
hep-th math-ph math.GT math.MP math.NT
|
http://creativecommons.org/licenses/by/4.0/
|
$\widehat{Z}$ invariants, rigorously defined for negative definite plumbed 3-manifolds, are expected--on physical grounds--to exist for every closed, oriented 3-manifold. Several prescriptions have been proposed to extend their definition to generic plumbings by reversing the orientation of a negative definite plumbing, thus turning it into a positive definite one. Two existing proposals are relevant for this paper: (i) the regularized $+1/r$-surgery conjecture combined with the false-mock modular conjecture, and (ii) a construction based on resurgence and a false theta function duality. In this note, we compare these proposals on the class of Brieskorn homology spheres $\Sigma\left(s,t,rst\pm1\right)$ and find that they are incompatible in general. Our diagnostic is the effective central charge, $c_{\text{eff}}$, which governs the asymptotic growth of coefficients of $\widehat{Z}$. First, we prove that the upper bound on $c_{\text{eff}}$ from prescription (i) is governed by the Ramanujan theta function, which regularizes the surgery formula. Second, we develop numerical and modular tools that deliver the lower bounds as well as exact values via mixed mock-modular analysis. Complementing this, we also study $c_{\text{eff}}$ for negative definite plumbed 3-manifolds which allow for a better comparison of pairs of 3-manifolds related by orientation reversal. As a result, we find that for some Brieskorn spheres the surgery and false-mock prescriptions violate the expected relation between $c_{\text{eff}}$, Chern-Simons invariants and non-abelian flat connections. These findings underscore $\widehat{Z}$ as a sensitive probe of the "positive side" of $\widehat{Z}$-theory.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:00 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Harichurn",
"Shimal",
""
],
[
"Jagadale",
"Mrunmay",
""
],
[
"Noshchenko",
"Dmitry",
""
],
[
"Passaro",
"Davide",
""
]
] |
2508.10088
|
Djuna Croon
|
Djuna Croon, Jeremy Sakstein, and Davide Gerosa
|
Can stellar physics explain GW231123?
|
4 pages, 2 figures
| null | null |
IPPP/25/52
|
astro-ph.HE astro-ph.GA astro-ph.SR gr-qc
|
http://creativecommons.org/licenses/by/4.0/
|
The gravitational wave event GW231123 detected by the LIGO interferometers during their fourth observing run features two black holes with source-frame masses of $137^{+22}_{-17} M_\odot$ and $103^{+20}_{-52} M_\odot $ -- well within or above the pair-instability black hole mass gap predicted by standard stellar evolution theory. Both black holes are also inferred to be rapidly spinning ($\chi_1 \simeq 0.9$, $\chi_2 \simeq 0.8$). The primary object in GW231123 is the heaviest stellar mass black hole detected to date, which, together with its extreme rotation, raises questions about its astrophysical origin. Accounting for the unusually large spin of $\sim 0.9$ with hierarchical mergers requires some degree of fine tuning. We investigate whether such a massive, highly spinning object could plausibly form from the collapse of a single rotating massive star. We simulate stars with an initial core mass of $160 \rm M_\odot$ -- sufficient to produce BH masses at the upper edge of the 90% credible interval for $m_1$ in GW231123 -- across a range of rotation rates and $^{12}\mathrm{C}(\alpha,\gamma)^{16}\mathrm{O}$ reaction rates. We find that: (i) rotation shifts the pair-instability mass gap to higher masses, introducing a significant ingredient that correlates masses and spins in gravitational wave predictions; and (ii) highly spinning BHs with masses $\gtrsim 150 \rm M_\odot$ can form above the mass gap, implying that stellar evolution alone is sufficient to explain GW231123. Our results suggest that the primary object of GW231123 may be the first directly observed black hole that formed via direct core collapse following the photodisintegration instability.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:00 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Croon",
"Djuna",
""
],
[
"Sakstein",
"Jeremy",
""
],
[
"Gerosa",
"Davide",
""
]
] |
2508.10089
|
Benjamin Davis PhD
|
Benjamin L. Davis, Mohamad Ali-Dib, Yujia Zheng, Zehao Jin, Kun Zhang, and Andrea Valerio Macci\`o
|
Causal evidence for the primordiality of colours in trans-Neptunian objects
|
9 pages, 5 figures, and 1 table; accepted by the Monthly Notices of the Royal Astronomical Society Letters
| null | null | null |
astro-ph.EP
|
http://creativecommons.org/licenses/by/4.0/
|
The origins of the colours of Trans-Neptunian Objects (TNOs) represent a crucial unresolved question, central to understanding the history of our Solar System. Recent observational surveys revealed correlations between the eccentricity and inclination of TNOs, and their colours. This rekindled the long-standing debate on whether these colours reflect the conditions of TNO formation or their subsequent evolution. We address this question using a model-agnostic, data-driven approach that unanimously converges to a common causal graph from the analysis of two different datasets, each from two different conditional independence test methods. For evaluation, we demonstrate how our model is consistent with the currently-accepted paradigms of TNOs' dynamical histories, without involving any orbital modelling or physics-based assumptions. Our causal model (with no knowledge of the existence of Neptune) predicts the need for an unknown confounding variable, consistent with Neptune's effects. The model predicts that the colour of TNOs is the root cause of their inclination distribution, rather than the other way around. This strongly suggests that the colours of TNOs reflect an underlying dynamical property, most likely their formation location. Our model excludes formation scenarios that invoke substantial colour modification by subsequent evolution. We conclude that the colours of TNOs are predominantly primordial.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:00 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Davis",
"Benjamin L.",
""
],
[
"Ali-Dib",
"Mohamad",
""
],
[
"Zheng",
"Yujia",
""
],
[
"Jin",
"Zehao",
""
],
[
"Zhang",
"Kun",
""
],
[
"Macciò",
"Andrea Valerio",
""
]
] |
2508.10090
|
Benoit Assi
|
Beno\^it Assi, Anthony Grebe and Michael Wagman
|
Baryon-baryon, meson-meson, and meson-baryon interactions in nonrelativistic QCD
|
23 pages, 6 figures
| null | null |
FERMILAB-PUB-25-0423-T
|
hep-ph hep-lat hep-th nucl-th
|
http://creativecommons.org/licenses/by/4.0/
|
Van der Waals potentials describing interactions between color-singlet mesons and/or baryons vanish at leading order in potential nonrelativistic quantum chromodynamics (pNRQCD). This result and constraints from Gauss's law are used to prove that weakly-coupled pNRQCD van der Waals potentials in generic non-Abelian gauge theories with only heavy quarks are too weak to form bound states whose color state is a product of color-singlets. Quantum Monte Carlo calculations of four, five, and six quarks with equal masses provide numerical evidence that exotic color configurations are higher energy than products of color-singlet hadrons, suggesting that equal-mass fully-heavy tetraquark, pentaquark, and hexaquark bound states do not exist at next-to-leading order in pNRQCD and at all orders in QCD-like theories in which all quark masses are asymptotically large. Mechanisms for generating hadron-hadron bound states are identified, which necessarily involve large quark-mass hierarchies, relativistic effects arising from the presence of sufficiently light quarks, or nonperturbative effects outside the scope of weakly-coupled pNRQCD.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:00 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Assi",
"Benoît",
""
],
[
"Grebe",
"Anthony",
""
],
[
"Wagman",
"Michael",
""
]
] |
2508.10091
|
Gaurang Ramakant Kane
|
Prateek Agrawal, Gaurang Ramakant Kane, Vazha Loladze, and John March-Russell
|
A Prediction for Maximum Supercooling in SU(N) Confinement Transition
|
6 pages, 4 figures
| null | null | null |
hep-ph hep-lat hep-th
|
http://creativecommons.org/licenses/by/4.0/
|
The thermal confinement phase transition (PT) in $SU(N)$ Yang-Mills theory is first-order for $N\geq 3$, with bounce action scaling as $N^2$. Remarkably, lattice data for the action include a small coefficient whose presence likely strongly alters the PT dynamics. We give evidence, utilizing insights from softly-broken SUSY YM models, that the small coefficient originates from a deconfined phase instability just below the critical temperature. We predict the maximum achievable supercooling in $SU(N)$ theories to be a few percent, which can be tested on the lattice. We briefly discuss the potentially significant suppression of the associated cosmological gravitational wave signals.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:01 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Agrawal",
"Prateek",
""
],
[
"Kane",
"Gaurang Ramakant",
""
],
[
"Loladze",
"Vazha",
""
],
[
"March-Russell",
"John",
""
]
] |
2508.10092
|
Yayaati Chachan
|
Yayaati Chachan, Joshua Lothringer, Julie Inglis, Hayley Beltz, Heather A. Knutson, Jessica Spake, Bjorn Benneke, Ian Wong, Zafar Rustamkulov, David Sing, Katherine A. Bennett
|
Strong NUV Refractory Absorption and Dissociated Water in the Hubble Transmission Spectrum of the Ultra Hot Jupiter KELT-20 b
|
28 pages, 18 figures, 6 tables, accepted for publication in AJ, comments welcome
| null | null | null |
astro-ph.EP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Ultra hot Jupiters (UHJs) present a promising pathway for drawing a link between a planet's composition and formation history. They retain both refractory and volatiles species in gas phase in their atmospheres, which allows us to place unique constraints on their building blocks. Here, we present the 0.2 - 1.7 $\mu$m transmission spectrum of KELT-20 b/MASCARA-2 b taken with the Hubble Space Telescope (HST). Unlike other UHJs around early-type stars, KELT-20 b's orbit is well aligned with its host star's spin axis and we test whether its distinct dynamical configuration is reflected in its composition. We observe a tremendous rise (>10 scale heights) in the planet's transit depth at the near-UV wavelengths, akin to that observed for WASP-178 b and WASP-121 b, and a muted water absorption feature in the near-IR. Our retrievals indicate that the large NUV depth is driven by Fe II and/or SiO and that the water is mostly thermally dissociated. Assuming equilibrium chemistry, we obtain constraints on Z/H and O/H that indicate accretion of volatile-rich solids and/or gas. Both our low resolution spectrum and the refractory elemental ratios from Gandhi et al. 2023 suggest that nightside condensation and rainout are limited to only the most refractory species in the planet's atmosphere. Within the precision limits of the HST spectra, no strong evidence for limb asymmetry is detected. We contextualize this lack of asymmetry by comparing to predictions from general circulation models with and without the effects of kinematic magnetohydrodynamics. Lastly, we find no major differences in the HST transmission spectra of KELT-20 b, WASP-178, and WASP-121 b despite their different dynamical configurations.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:01 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Chachan",
"Yayaati",
""
],
[
"Lothringer",
"Joshua",
""
],
[
"Inglis",
"Julie",
""
],
[
"Beltz",
"Hayley",
""
],
[
"Knutson",
"Heather A.",
""
],
[
"Spake",
"Jessica",
""
],
[
"Benneke",
"Bjorn",
""
],
[
"Wong",
"Ian",
""
],
[
"Rustamkulov",
"Zafar",
""
],
[
"Sing",
"David",
""
],
[
"Bennett",
"Katherine A.",
""
]
] |
2508.10093
|
Vyshnav Mohan
|
Vyshnav Mohan and Watse Sybesma
|
De Sitter Complexity Grows Linearly in the Static Patch
|
6 pages and 3 figures
| null | null | null |
hep-th gr-qc
|
http://creativecommons.org/licenses/by/4.0/
|
The observable universe has undergone periods of expansion that are well approximated by de Sitter (dS) space. Still lacking is a quantum mechanical description of dS, both globally and when restricted to the static patch. We develop a novel prescription for computing holographic complexity in the dS static patch to determine its microscopic features. Specifically, we propose that the natural candidate for dS complexity is the volume of extremal timelike surfaces restricted to the static patch, anchored to the cosmological horizon or an observer worldline. Our anchoring prescription provides a clear definition of a reference state, overcoming a common ambiguity in prior definitions of de Sitter holographic complexity. The late-time growth of our complexity functional is linear and proportional to the number of degrees of freedom associated to the cosmological horizon, and therefore does not exhibit hyperfast growth. Our results imply the dS static patch is characterized by a quantum mechanical system, with a finite dimensional Hilbert space whose evolution is governed by a chaotic Hamiltonian.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:01 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Mohan",
"Vyshnav",
""
],
[
"Sybesma",
"Watse",
""
]
] |
2508.10094
|
Benedikt Schosser
|
Benedikt Schosser, Caroline Heneka, Bj\"orn Malte Sch\"afer
|
Starobinsky in Stereo: SKA-CMB Synergy in SBI
|
22 pages, 9 figures
| null | null | null |
astro-ph.CO astro-ph.IM physics.data-an
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Modern machine learning techniques can unlock the vast cosmological information encoded in forthcoming Square Kilometre Array (SKA) observations. We show that tomographic 21 cm data from the reionisation era can yield stringent tests of inflationary models - here illustrated with Starobinsky $R+R^2$ inflation. Using a simulation-based inference (SBI) framework, we compare neural summaries (convolutional network and vision transformer) with a traditional power spectrum summary and perform a fully joint SBI analysis combining 21 cm data with data of the cosmic microwave background (CMB). Forecasts based on realistic mock observations indicate that SKA alone will achieve constraints competitive with Planck, and that the combined SKA + CMB dataset will tighten bounds on both inflationary and $\Lambda\mathrm{CDM}$ parameters considerably while improving precision on key astrophysical quantities.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:02 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Schosser",
"Benedikt",
""
],
[
"Heneka",
"Caroline",
""
],
[
"Schäfer",
"Björn Malte",
""
]
] |
2508.10095
|
Pengfei Li
|
Pengfei Li and Zheng Zheng
|
Emulating extended Lyman-alpha haloes around star-forming galaxies
|
16 pages, 15 figures, accepted by MNRAS
| null | null | null |
astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
Extended Ly$\alpha$ emission is commonly observed around star-forming galaxies, opening a window for probing the neutral hydrogen gas in the circumgalactic medium (CGM). In this paper, we develop a prescription of spherically symmetric CGM gas properties and build emulators to model circularly-averaged surface brightness (SB) profiles of the extended Ly$\alpha$ emission. With CGM gas properties parametrized by the density, velocity, and temperature profiles, a self-shielding calculation is carried out to obtain the neutral gas distribution with ionizing photons from the ultraviolet (UV) background and star formation in the galaxy. Our calculation reveals three types of systems with distinct neutral gas distribution: non-shielded systems with the CGM being highly ionized across all radii, shielded systems with a neutral gas shell shielding the UV background, and transitional systems in between. Ly$\alpha$ SB profiles are obtained through Ly$\alpha$ radiative transfer (RT) simulations, performed for the CGM models with three kinds of Ly$\alpha$ sources: the star formation from central and satellite galaxies, and the recombination in the CGM. We build emulators to efficiently predict Ly$\alpha$ SB profiles for given model parameters and Ly$\alpha$ sources, based on Gaussian process regression. After being trained with only 180 RT simulations for each Ly$\alpha$ source, the emulators reach an overall accuracy at the level of $\sim 20$ per cent. By applying the emulators to fit mock Ly$\alpha$ SB profiles constructed from our model, we find a reasonable recovery of model parameters, indicating the potential of extracting physical information of the CGM and galaxies from the observed extended Ly$\alpha$ emission.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:02 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Li",
"Pengfei",
""
],
[
"Zheng",
"Zheng",
""
]
] |
2508.10096
|
Aaron Sander
|
Aaron Sander, Maximilian Fr\"ohlich, Mazen Ali, Martin Eigel, Jens Eisert, Michael Hinterm\"uller, Christian B. Mendl, Richard M. Milbradt, Robert Wille
|
Quantum circuit simulation with a local time-dependent variational principle
|
13 pages, 2 figures
| null | null | null |
quant-ph cond-mat.other
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Classical simulations of quantum circuits are vital for assessing potential quantum advantage and benchmarking devices, yet they require sophisticated methods to avoid the exponential growth of resources. Tensor network approaches, in particular matrix product states (MPS) combined with the time-evolving block decimation (TEBD) algorithm, currently dominate large-scale circuit simulations. These methods scale efficiently when entanglement is limited but suffer rapid bond dimension growth with increasing entanglement and handle long-range gates via costly SWAP insertions. Motivated by the success of the time-dependent variational principle (TDVP) in many-body physics, we reinterpret quantum circuits as a series of discrete time evolutions, using gate generators to construct an MPS-based circuit simulation via a local TDVP formulation. This addresses TEBD's key limitations by (1) naturally accommodating long-range gates and (2) optimally representing states on the MPS manifold. By diffusing entanglement more globally, the method suppresses local bond growth and reduces memory and runtime costs. We benchmark the approach on five 49-qubit circuits: three Hamiltonian circuits (1D open and periodic Heisenberg, 2D 7x7 Ising) and two algorithmic ones (quantum approximate optimization, hardware-efficient ansatz). Across all cases, our method yields substantial resource reductions over standard tools, establishing a new state-of-the-art for circuit simulation and enabling advances across quantum computing, condensed matter, and beyond.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:02 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Sander",
"Aaron",
""
],
[
"Fröhlich",
"Maximilian",
""
],
[
"Ali",
"Mazen",
""
],
[
"Eigel",
"Martin",
""
],
[
"Eisert",
"Jens",
""
],
[
"Hintermüller",
"Michael",
""
],
[
"Mendl",
"Christian B.",
""
],
[
"Milbradt",
"Richard M.",
""
],
[
"Wille",
"Robert",
""
]
] |
2508.10097
|
Michael Shull
|
J. Michael Shull (1), Rongmon Bordoloi (2), Charles W. Danforth (1) ((1) University of Colorado, (2) North Carolina State University)
|
Quasar Spectral Energy Distributions in the Rest-Frame EUV: Hubble-COS Spectra of Two Ultra-luminous Quasars
|
Accepted to the Astrophysical Journal, 12 pages, 3 figures, 3 tables
| null | null | null |
astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
Using the Cosmic Origins Spectrograph (COS) aboard the Hubble Space Telescope with both far-UV (FUV) and near-UV (NUV) gratings, we measure the ionizing spectra of two bright, intermediate-redshift quasars in their rest-frame extreme ultraviolet (EUV). The availability of both NUV and FUV spectra allows us to define the quasar continuum and correct for strong Lyman-limit systems (LLS) that fall in the gap between the FUV and optical. Each AGN has a prominent LLS, but the flux recovery shortward of their Lyman edges allows us to fit and restore the true AGN continuum. In the EUV (450-912 A) these AGN have flux distributions, $F_{\nu} \propto \nu^{-\alpha_{\nu}}$, with spectral indices $\alpha_{\nu} = 1.11\pm0.22$ (SBS 1010+535, $z_{\rm AGN} = 1.5086$) and $\alpha_{\nu} = 0.98\pm0.22$ (HS 0747+4259, $z_{\rm AGN} = 1.9006$), both considerably harder than the mean index, $\alpha_{\nu} = 1.41\pm0.15$, in a COS composite spectrum of 159 UV-bright AGN. These two AGN are outliers in the index distribution, perhaps resulting from their extremely high UV luminosity ($10^{48}~{\rm erg~s}^{-1}$), estimated black-hole masses (0.5-1)$\times10^{10} M_{\odot}$, and effects on the inner accretion disk and Comptonized winds.
|
[
{
"version": "v1",
"created": "Wed, 13 Aug 2025 18:00:02 GMT"
}
] | 2025-08-15T00:00:00 |
[
[
"Shull",
"J. Michael",
"",
"University of Colorado"
],
[
"Bordoloi",
"Rongmon",
"",
"North Carolina State University"
],
[
"Danforth",
"Charles W.",
"",
"University of Colorado"
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.