LLM 관련 주요 논문 - 2026-01-23

1. Structured Hints for Sample-Efficient Lean Theorem Proving

Authors: Zachary Burton
URL: https://arxiv.org/abs/2601.16172
Abstract:

State-of-the-art neural theorem provers like DeepSeek-Prover-V1.5 combine large language models with reinforcement learning, achieving impressive results through sophisticated training. We ask: do these highly-trained models still benefit from simple structural guidance at inference time? We evaluate a lightweight intervention – a fixed prompt schedule over 15 common tactic skeletons – on the miniF2F benchmark. This simple approach yields 21.7% pass@16 compared to 15.2% for standard sampling from the same model, a 43% relative improvement using the same number of samples (k=16) and same maximum generation length (1024 tokens). Our results suggest that even capable RL-trained provers underutilize structural priors available in the tactic language, and that simple inference-time guidance remains a cheap, complementary boost.

2. LLM Prompt Evaluation for Educational Applications

Authors: Langdon Holmes , Adam Coscia , Scott Crossley , Joon Suh Choi , Wesley Morris
URL: https://arxiv.org/abs/2601.16134
Abstract:

As large language models (LLMs) become increasingly common in educational applications, there is a growing need for evidence-based methods to design and evaluate LLM prompts that produce personalized and pedagogically aligned out-puts. This study presents a generalizable, systematic approach for evaluating prompts, demonstrated through an analysis of LLM-generated follow-up questions in a structured dialogue activity. Six prompt templates were designed and tested. The templates incorporated established prompt engineering patterns, with each prompt emphasizing distinct pedagogical strategies. The prompt templates were compared through a tournament-style evaluation framework that can be adapted for other educational applications. The tournament employed the Glicko2 rating system with eight judges evaluating question pairs across three dimensions: format, dialogue support, and appropriateness for learners. Data was sourced from 120 authentic user interactions across three distinct educational deployments. Results showed that a single prompt related to strategic reading out-performed other templates with win probabilities ranging from 81% to 100% in pairwise comparisons. This prompt combined persona and context manager pat-terns and was designed to support metacognitive learning strategies such as self-directed learning. The methodology showcases how educational technology re- searchers can systematically evaluate and improve prompt designs, moving beyond ad-hoc prompt engineering toward evidence-based prompt development for educational applications.

3. Multimodal Climate Disinformation Detection: Integrating Vision-Language Models with External Knowledge Sources

Authors: Marzieh Adeli Shamsabad , Hamed Ghodrati
URL: https://arxiv.org/abs/2601.16108
Abstract:

Climate disinformation has become a major challenge in today digital world, especially with the rise of misleading images and videos shared widely on social media. These false claims are often convincing and difficult to detect, which can delay actions on climate change. While vision-language models (VLMs) have been used to identify visual disinformation, they rely only on the knowledge available at the time of training. This limits their ability to reason about recent events or updates. The main goal of this paper is to overcome that limitation by combining VLMs with external knowledge. By retrieving up-to-date information such as reverse image results, online fact-checks, and trusted expert content, the system can better assess whether an image and its claim are accurate, misleading, false, or unverifiable. This approach improves the model ability to handle real-world climate disinformation and supports efforts to protect public understanding of science in a rapidly changing information landscape.

4. Controlling Long-Horizon Behavior in Language Model Agents with Explicit State Dynamics

Authors: Sukesh Subaharan
URL: https://arxiv.org/abs/2601.16087
Abstract:

Abstract not available

5. Grounding Large Language Models in Reaction Knowledge Graphs for Synthesis Retrieval

Authors: Olga Bunkova , Lorenzo Di Fruscia , Sophia Rupprecht , Artur M. Schweidtmann , Marcel J.T. Reinders , Jana M. Weber
URL: https://arxiv.org/abs/2601.16038
Abstract:

Large Language Models (LLMs) can aid synthesis planning in chemistry, but standard prompting methods often yield hallucinated or outdated suggestions. We study LLM interactions with a reaction knowledge graph by casting reaction path retrieval as a Text2Cypher (natural language to graph query) generation problem, and define single- and multi-step retrieval tasks. We compare zero-shot prompting to one-shot variants using static, random, and embedding-based exemplar selection, and assess a checklist-driven validator/corrector loop. To evaluate our framework, we consider query validity and retrieval accuracy. We find that one-shot prompting with aligned exemplars consistently performs best. Our checklist-style self-correction loop mainly improves executability in zero-shot settings and offers limited additional retrieval gains once a good exemplar is present. We provide a reproducible Text2Cypher evaluation setup to facilitate further work on KG-grounded LLMs for synthesis planning. Code is available at this https URL .

6. Deja Vu in Plots: Leveraging Cross-Session Evidence with Retrieval-Augmented LLMs for Live Streaming Risk Assessment

Authors: Yiran Qiao , Xiang Ao , Jing Chen , Yang Liu , Qiwei Zhong , Qing He
URL: https://arxiv.org/abs/2601.16027
Abstract:

The rise of live streaming has transformed online interaction, enabling massive real-time engagement but also exposing platforms to complex risks such as scams and coordinated malicious behaviors. Detecting these risks is challenging because harmful actions often accumulate gradually and recur across seemingly unrelated streams. To address this, we propose CS-VAR (Cross-Session Evidence-Aware Retrieval-Augmented Detector) for live streaming risk assessment. In CS-VAR, a lightweight, domain-specific model performs fast session-level risk inference, guided during training by a Large Language Model (LLM) that reasons over retrieved cross-session behavioral evidence and transfers its local-to-global insights to the small model. This design enables the small model to recognize recurring patterns across streams, perform structured risk assessment, and maintain efficiency for real-time deployment. Extensive offline experiments on large-scale industrial datasets, combined with online validation, demonstrate the state-of-the-art performance of CS-VAR. Furthermore, CS-VAR provides interpretable, localized signals that effectively empower real-world moderation for live streaming.

7. ErrorMap and ErrorAtlas: Charting the Failure Landscape of Large Language Models

Authors: Shir Ashury-Tahan , Yifan Mai , Elron Bandel , Michal Shmueli-Scheuer , Leshem Choshen
URL: https://arxiv.org/abs/2601.15812
Abstract:

Large Language Models (LLM) benchmarks tell us when models fail, but not why they fail. A wrong answer on a reasoning dataset may stem from formatting issues, calculation errors, or dataset noise rather than weak reasoning. Without disentangling such causes, benchmarks remain incomplete and cannot reliably guide model improvement. We introduce ErrorMap, the first method to chart the sources of LLM failure. It extracts a model’s unique “failure signature”, clarifies what benchmarks measure, and broadens error identification to reduce blind spots. This helps developers debug models, aligns benchmark goals with outcomes, and supports informed model selection. ErrorMap works on any model or dataset with the same logic. Applying our method to 35 datasets and 83 models we generate ErrorAtlas, a taxonomy of model errors, revealing recurring failure patterns. ErrorAtlas highlights error types that are currently underexplored in LLM research, such as omissions of required details in the output and question misinterpretation. By shifting focus from where models succeed to why they fail, ErrorMap and ErrorAtlas enable advanced evaluation - one that exposes hidden weaknesses and directs progress. Unlike success, typically measured by task-level metrics, our approach introduces a deeper evaluation layer that can be applied globally across models and tasks, offering richer insights into model behavior and limitations. We make the taxonomy and code publicly available with plans to periodically update ErrorAtlas as new benchmarks and models emerge.

8. Inference-Time Scaling of Verification: Self-Evolving Deep Research Agents via Test-Time Rubric-Guided Verification

Authors: Yuxuan Wan , Tianqing Fang , Zaitang Li , Yintong Huo , Wenxuan Wang , Haitao Mi , Dong Yu , Michael R. Lyu
URL: https://arxiv.org/abs/2601.15808
Abstract:

Recent advances in Deep Research Agents (DRAs) are transforming automated knowledge discovery and problem-solving. While the majority of existing efforts focus on enhancing policy capabilities via post-training, we propose an alternative paradigm: self-evolving the agent’s ability by iteratively verifying the policy model’s outputs, guided by meticulously crafted rubrics. This approach gives rise to the inference-time scaling of verification, wherein an agent self-improves by evaluating its generated answers to produce iterative feedback and refinements. We derive the rubrics based on an automatically constructed DRA Failure Taxonomy, which systematically classifies agent failures into five major categories and thirteen sub-categories. We present DeepVerifier, a rubrics-based outcome reward verifier that leverages the asymmetry of verification and outperforms vanilla agent-as-judge and LLM judge baselines by 12%-48% in meta-evaluation F1 score. To enable practical self-evolution, DeepVerifier integrates as a plug-and-play module during test-time inference. The verifier produces detailed rubric-based feedback, which is fed back to the agent for iterative bootstrapping, refining responses without additional training. This test-time scaling delivers 8%-11% accuracy gains on challenging subsets of GAIA and XBench-DeepResearch when powered by capable closed-source LLMs. Finally, to support open-source advancement, we release DeepVerifier-4K, a curated supervised fine-tuning dataset of 4,646 high-quality agent steps focused on DRA verification. These examples emphasize reflection and self-critique, enabling open models to develop robust verification capabilities.

9. VitalDiagnosis: AI-Driven Ecosystem for 24/7 Vital Monitoring and Chronic Disease Management

Authors: Zhikai Xue , Tianqianjin Lin , Pengwei Yan , Ruichun Wang , Yuxin Liu , Zhuoren Jiang , Xiaozhong Liu
URL: https://arxiv.org/abs/2601.15798
Abstract:

Chronic diseases have become the leading cause of death worldwide, a challenge intensified by strained medical resources and an aging population. Individually, patients often struggle to interpret early signs of deterioration or maintain adherence to care plans. In this paper, we introduce VitalDiagnosis, an LLM-driven ecosystem designed to shift chronic disease management from passive monitoring to proactive, interactive engagement. By integrating continuous data from wearable devices with the reasoning capabilities of LLMs, the system addresses both acute health anomalies and routine adherence. It analyzes triggers through context-aware inquiries, produces provisional insights within a collaborative patient-clinician workflow, and offers personalized guidance. This approach aims to promote a more proactive and cooperative care paradigm, with the potential to enhance patient self-management and reduce avoidable clinical workload.

10. Agentic Confidence Calibration

Authors: Jiaxin Zhang , Caiming Xiong , Chien-Sheng Wu
URL: https://arxiv.org/abs/2601.15778
Abstract:

AI agents are rapidly advancing from passive language models to autonomous systems executing complex, multi-step tasks. Yet their overconfidence in failure remains a fundamental barrier to deployment in high-stakes settings. Existing calibration methods, built for static single-turn outputs, cannot address the unique challenges of agentic systems, such as compounding errors along trajectories, uncertainty from external tools, and opaque failure modes. To address these challenges, we introduce, for the first time, the problem of Agentic Confidence Calibration and propose Holistic Trajectory Calibration (HTC), a novel diagnostic framework that extracts rich process-level features ranging from macro dynamics to micro stability across an agent’s entire trajectory. Powered by a simple, interpretable model, HTC consistently surpasses strong baselines in both calibration and discrimination, across eight benchmarks, multiple LLMs, and diverse agent frameworks. Beyond performance, HTC delivers three essential advances: it provides interpretability by revealing the signals behind failure, enables transferability by applying across domains without retraining, and achieves generalization through a General Agent Calibrator (GAC) that achieves the best calibration (lowest ECE) on the out-of-domain GAIA benchmark. Together, these contributions establish a new process-centric paradigm for confidence calibration, providing a framework for diagnosing and enhancing the reliability of AI agents.

11. Tabular Incremental Inference

Authors: Xinda Chen , Xing Zhen , Hanyu Zhang , Weimin Tan , Bo Yan
URL: https://arxiv.org/abs/2601.15751
Abstract:

Tabular data is a fundamental form of data structure. The evolution of table analysis tools reflects humanity’s continuous progress in data acquisition, management, and processing. The dynamic changes in table columns arise from technological advancements, changing needs, data integration, etc. However, the standard process of training AI models on tables with fixed columns and then performing inference is not suitable for handling dynamically changed tables. Therefore, new methods are needed for efficiently handling such tables in an unsupervised manner. In this paper, we introduce a new task, Tabular Incremental Inference (TabII), which aims to enable trained models to incorporate new columns during the inference stage, enhancing the practicality of AI models in scenarios where tables are dynamically changed. Furthermore, we demonstrate that this new task can be framed as an optimization problem based on the information bottleneck theory, which emphasizes that the key to an ideal tabular incremental inference approach lies in minimizing mutual information between tabular data and representation while maximizing between representation and task labels. Under this guidance, we design a TabII method with Large Language Model placeholders and Pretrained TabAdapter to provide external knowledge and Incremental Sample Condensation blocks to condense the task-relevant information given by incremental column attributes. Experimental results across eight public datasets show that TabII effectively utilizes incremental attributes, achieving state-of-the-art performance.

12. AgentSM: Semantic Memory for Agentic Text-to-SQL

Authors: Asim Biswal , Chuan Lei , Xiao Qin , Aodong Li , Balakrishnan Narayanaswamy , Tim Kraska
URL: https://arxiv.org/abs/2601.15709
Abstract:

Recent advances in LLM-based Text-to-SQL have achieved remarkable gains on public benchmarks such as BIRD and Spider. Yet, these systems struggle to scale in realistic enterprise settings with large, complex schemas, diverse SQL dialects, and expensive multi-step reasoning. Emerging agentic approaches show potential for adaptive reasoning but often suffer from inefficiency and instability-repeating interactions with databases, producing inconsistent outputs, and occasionally failing to generate valid answers. To address these challenges, we introduce Agent Semantic Memory (AgentSM), an agentic framework for Text-to-SQL that builds and leverages interpretable semantic memory. Instead of relying on raw scratchpads or vector retrieval, AgentSM captures prior execution traces-or synthesizes curated ones-as structured programs that directly guide future reasoning. This design enables systematic reuse of reasoning paths, which allows agents to scale to larger schemas, more complex questions, and longer trajectories efficiently and reliably. Compared to state-of-the-art systems, AgentSM achieves higher efficiency by reducing average token usage and trajectory length by 25% and 35%, respectively, on the Spider 2.0 benchmark. It also improves execution accuracy, reaching a state-of-the-art accuracy of 44.8% on the Spider 2.0 Lite benchmark.

13. Improving Methodologies for LLM Evaluations Across Global Languages

Authors: Akriti Vij , Benjamin Chua , Darshini Ramiah , En Qi Ng , Mahran Morsidi , Naga Nikshith Gangarapu , Sharmini Johnson , Vanessa Wilfred , Vikneswaran Kumaran , Wan Sie Lee , Wenzhuo Yang , Yongsen Zheng , Bill Black , Boming Xia , Frank Sun , Hao Zhang , Qinghua Lu , Suyu Ma , Yue Liu , Chi-kiu Lo , Fatemeh Azadi , Isar Nejadgholi , Sowmya Vajjala , Agnes Delaborde , Nicolas Rolin , Tom Seimandi , Akiko Murakami , Haruto Ishi , Satoshi Sekine , Takayuki Semitsu , Tasuku Sasaki , Angela Kinuthia , Jean Wangari , Michael Michie , Stephanie Kasaon , Hankyul Baek , Jaewon Noh , Kihyuk Nam , Sang Seo , Sungpil Shin , Taewhi Lee , Yongsu Kim , Daisy Newbold-Harrop , Jessica Wang , Mahmoud Ghanem , Vy Hong
URL: https://arxiv.org/abs/2601.15706
Abstract:

As frontier AI models are deployed globally, it is essential that their behaviour remains safe and reliable across diverse linguistic and cultural contexts. To examine how current model safeguards hold up in such settings, participants from the International Network for Advanced AI Measurement, Evaluation and Science, including representatives from Singapore, Japan, Australia, Canada, the EU, France, Kenya, South Korea and the UK conducted a joint multilingual evaluation exercise. Led by Singapore AISI, two open-weight models were tested across ten languages spanning high and low resourced groups: Cantonese English, Farsi, French, Japanese, Korean, Kiswahili, Malay, Mandarin Chinese and Telugu. Over 6,000 newly translated prompts were evaluated across five harm categories (privacy, non-violent crime, violent crime, intellectual property and jailbreak robustness), using both LLM-as-a-judge and human annotation. The exercise shows how safety behaviours can vary across languages. These include differences in safeguard robustness across languages and harm types and variation in evaluator reliability (LLM-as-judge vs. human review). Further, it also generated methodological insights for improving multilingual safety evaluations, such as the need for culturally contextualised translations, stress-tested evaluator prompts and clearer human annotation guidelines. This work represents an initial step toward a shared framework for multilingual safety testing of advanced AI systems and calls for continued collaboration with the wider research community and industry.

14. From Passive Metric to Active Signal: The Evolving Role of Uncertainty Quantification in Large Language Models

Authors: Jiaxin Zhang , Wendi Cui , Zhuohang Li , Lifu Huang , Bradley Malin , Caiming Xiong , Chien-Sheng Wu
URL: https://arxiv.org/abs/2601.15690
Abstract:

While Large Language Models (LLMs) show remarkable capabilities, their unreliability remains a critical barrier to deployment in high-stakes domains. This survey charts a functional evolution in addressing this challenge: the evolution of uncertainty from a passive diagnostic metric to an active control signal guiding real-time model behavior. We demonstrate how uncertainty is leveraged as an active control signal across three frontiers: in \textbf{advanced reasoning} to optimize computation and trigger self-correction; in \textbf{autonomous agents} to govern metacognitive decisions about tool use and information seeking; and in \textbf{reinforcement learning} to mitigate reward hacking and enable self-improvement via intrinsic rewards. By grounding these advancements in emerging theoretical frameworks like Bayesian methods and Conformal Prediction, we provide a unified perspective on this transformative trend. This survey provides a comprehensive overview, critical analysis, and practical design patterns, arguing that mastering the new trend of uncertainty is essential for building the next generation of scalable, reliable, and trustworthy AI.

15. Predictive Coding and Information Bottleneck for Hallucination Detection in Large Language Models

Authors: Manish Bhatt
URL: https://arxiv.org/abs/2601.15652
Abstract:

Hallucinations in Large Language Models (LLMs) – generations that are plausible but factually unfaithful – remain a critical barrier to high-stakes deployment. Current detection methods typically rely on computationally expensive external retrieval loops or opaque black-box LLM judges requiring 70B+ parameters. In this work, we introduce [Model Name], a hybrid detection framework that combines neuroscience-inspired signal design with supervised machine learning. We extract interpretable signals grounded in Predictive Coding (quantifying surprise against internal priors) and the Information Bottleneck (measuring signal retention under perturbation). Through systematic ablation, we demonstrate three key enhancements: Entity-Focused Uptake (concentrating on high-value tokens), Context Adherence (measuring grounding strength), and Falsifiability Score (detecting confident but contradictory claims). Evaluating on HaluBench (n=200, perfectly balanced), our theory-guided baseline achieves 0.8017 AUROC. BASE supervised models reach 0.8274 AUROC, while IMPROVED features boost performance to 0.8669 AUROC (4.95% gain), demonstrating consistent improvements across architectures. This competitive performance is achieved while using 75x less training data than Lynx (200 vs 15,000 samples), 1000x faster inference (5ms vs 5s), and remaining fully interpretable. Crucially, we report a negative result: the Rationalization signal fails to distinguish hallucinations, suggesting that LLMs generate coherent reasoning for false premises (“Sycophancy”). This work demonstrates that domain knowledge encoded in signal architecture provides superior data efficiency compared to scaling LLM judges, achieving strong performance with lightweight (less than 1M parameter), explainable models suitable for production deployment.

16. CogToM: A Comprehensive Theory of Mind Benchmark inspired by Human Cognition for Large Language Models

Authors: Haibo Tong , Zeyang Yue , Feifei Zhao , Erliang Lin , Lu Jia , Ruolin Chen , Yinqian Sun , Qian Zhang , Yi Zeng
URL: https://arxiv.org/abs/2601.15628
Abstract:

Whether Large Language Models (LLMs) truly possess human-like Theory of Mind (ToM) capabilities has garnered increasing attention. However, existing benchmarks remain largely restricted to narrow paradigms like false belief tasks, failing to capture the full spectrum of human cognitive mechanisms. We introduce CogToM, a comprehensive, theoretically grounded benchmark comprising over 8000 bilingual instances across 46 paradigms, validated by 49 human annotator.A systematic evaluation of 22 representative models, including frontier models like GPT-5.1 and Qwen3-Max, reveals significant performance heterogeneities and highlights persistent bottlenecks in specific dimensions. Further analysis based on human cognitive patterns suggests potential divergences between LLM and human cognitive structures. CogToM offers a robust instrument and perspective for investigating the evolving cognitive boundaries of LLMs.

17. Autonomous Business System via Neuro-symbolic AI

Authors: Cecil Pang , Hiroki Sayama
URL: https://arxiv.org/abs/2601.15599
Abstract:

Current business environments require organizations to continuously reconfigure cross-functional processes, yet enterprise systems are still organized around siloed departments, rigid workflows, and hard-coded automation. Meanwhile large language models (LLMs) excel at interpreting natural language and unstructured data but lack deterministic, verifiable execution of complex business logic. To address this gap, here we introduce AUTOBUS, an Autonomous Business System that integrates LLM-based AI agents, predicate-logic programming, and business-semantics-centric enterprise data into a coherent neuro-symbolic AI architecture for orchestrating end-to-end business initiatives. AUTOBUS models an initiative as a network of tasks with explicit pre/post conditions, required data, evaluation rules, and API-level actions. Enterprise data is organized as a knowledge graph whose entities, relationships, and constraints are translated into logic facts and foundational rules, providing the semantic grounding for task reasoning. Core AI agents synthesize task instructions, enterprise semantics, and available tools into task-specific logic programs, which are executed by a logic engine that enforces constraints, coordinates auxiliary tools, and orchestrate execution of actions and outcomes. Humans define and maintain the semantics, policies and task instructions, curate tools, and supervise high-impact or ambiguous decisions, ensuring accountability and adaptability. We detail the AUTOBUS architecture, the anatomy of the AI agent generated logic programs, and the role of humans and auxiliary tools in the lifecycle of a business initiative.

18. TransportAgents: a multi-agents LLM framework for traffic accident severity prediction

Authors: Zhichao Yang , Jiashu He , Jinxuan Fan , Cirillo Cinzia
URL: https://arxiv.org/abs/2601.15519
Abstract:

Accurate prediction of traffic crash severity is critical for improving emergency response and public safety planning. Although recent large language models (LLMs) exhibit strong reasoning capabilities, their single-agent architectures often struggle with heterogeneous, domain-specific crash data and tend to generate biased or unstable predictions. To address these limitations, this paper proposes TransportAgents, a hybrid multi-agent framework that integrates category-specific LLM reasoning with a multilayer perceptron (MLP) integration module. Each specialized agent focuses on a particular subset of traffic information, such as demographics, environmental context, or incident details, to produce intermediate severity assessments that are subsequently fused into a unified prediction. Extensive experiments on two complementary U.S. datasets, the Consumer Product Safety Risk Management System (CPSRMS) and the National Electronic Injury Surveillance System (NEISS), demonstrate that TransportAgents consistently outperforms both traditional machine learning and advanced LLM-based baselines. Across three representative backbones, including closed-source models such as GPT-3.5 and GPT-4o, as well as open-source models such as LLaMA-3.3, the framework exhibits strong robustness, scalability, and cross-dataset generalizability. A supplementary distributional analysis further shows that TransportAgents produces more balanced and well-calibrated severity predictions than standard single-agent LLM approaches, highlighting its interpretability and reliability for safety-critical decision support applications.

19. Tracking the Limits of Knowledge Propagation: How LLMs Fail at Multi-Step Reasoning with Conflicting Knowledge

Authors: Yiyang Feng , Zeming Chen , Haotian Wu , Jiawei Zhou , Antoine Bosselut
URL: https://arxiv.org/abs/2601.15495
Abstract:

A common solution for mitigating outdated or incorrect information in Large Language Models (LLMs) is to provide updated facts in-context or through knowledge editing. However, these methods introduce knowledge conflicts when the knowledge update fails to overwrite the model’s parametric knowledge, which propagate to faulty reasoning. Current benchmarks for this problem, however, largely focus only on single knowledge updates and fact recall without evaluating how these updates affect downstream reasoning. In this work, we introduce TRACK (Testing Reasoning Amid Conflicting Knowledge), a new benchmark for studying how LLMs propagate new knowledge through multi-step reasoning when it conflicts with the model’s initial parametric knowledge. Spanning three reasoning-intensive scenarios (WIKI, CODE, and MATH), TRACK introduces multiple, realistic conflicts to mirror real-world complexity. Our results on TRACK reveal that providing updated facts to models for reasoning can worsen performance compared to providing no updated facts to a model, and that this performance degradation exacerbates as more updated facts are provided. We show this failure stems from both inability to faithfully integrate updated facts, but also flawed reasoning even when knowledge is integrated. TRACK provides a rigorous new benchmark to measure and guide future progress on propagating conflicting knowledge in multi-step reasoning.

20. Reliability by design: quantifying and eliminating fabrication risk in LLMs. From generative to consultative AI: a comparative analysis in the legal domain and lessons for high-stakes knowledge bases

Authors: Alex Dantart
URL: https://arxiv.org/abs/2601.15476
Abstract:

This paper examines how to make large language models reliable for high-stakes legal work by reducing hallucinations. It distinguishes three AI paradigms: (1) standalone generative models (“creative oracle”), (2) basic retrieval-augmented systems (“expert archivist”), and (3) an advanced, end-to-end optimized RAG system (“rigorous archivist”). The authors introduce two reliability metrics -False Citation Rate (FCR) and Fabricated Fact Rate (FFR)- and evaluate 2,700 judicial-style answers from 12 LLMs across 75 legal tasks using expert, double-blind review. Results show that standalone models are unsuitable for professional use (FCR above 30%), while basic RAG greatly reduces errors but still leaves notable misgrounding. Advanced RAG, using techniques such as embedding fine-tuning, re-ranking, and self-correction, reduces fabrication to negligible levels (below 0.2%). The study concludes that trustworthy legal AI requires rigor-focused, retrieval-based architectures emphasizing verification and traceability, and provides an evaluation framework applicable to other high-risk domains.

21. Not Your Typical Sycophant: The Elusive Nature of Sycophancy in Large Language Models

Authors: Shahar Ben Natan , Oren Tsur
URL: https://arxiv.org/abs/2601.15436
Abstract:

We propose a novel way to evaluate sycophancy of LLMs in a direct and neutral way, mitigating various forms of uncontrolled bias, noise, or manipulative language, deliberately injected to prompts in prior works. A key novelty in our approach is the use of LLM-as-a-judge, evaluation of sycophancy as a zero-sum game in a bet setting. Under this framework, sycophancy serves one individual (the user) while explicitly incurring cost on another. Comparing four leading models - Gemini 2.5 Pro, ChatGpt 4o, Mistral-Large-Instruct-2411, and Claude Sonnet 3.7 - we find that while all models exhibit sycophantic tendencies in the common setting, in which sycophancy is self-serving to the user and incurs no cost on others, Claude and Mistral exhibit “moral remorse” and over-compensate for their sycophancy in case it explicitly harms a third party. Additionally, we observed that all models are biased toward the answer proposed last. Crucially, we find that these two phenomena are not independent; sycophancy and recency bias interact to produce `constructive interference’ effect, where the tendency to agree with the user is exacerbated when the user’s opinion is presented last.

22. Beyond Prompting: Efficient and Robust Contextual Biasing for Speech LLMs via Logit-Space Integration (LOGIC)

Authors: Peidong Wang
URL: https://arxiv.org/abs/2601.15397
Abstract:

The rapid emergence of new entities – driven by cultural shifts, evolving trends, and personalized user data – poses a significant challenge for existing Speech Large Language Models (Speech LLMs). While these models excel at general conversational tasks, their static training knowledge limits their ability to recognize domain-specific terms such as contact names, playlists, or technical jargon. Existing solutions primarily rely on prompting, which suffers from poor scalability: as the entity list grows, prompting encounters context window limitations, increased inference latency, and the “lost-in-the-middle” phenomenon. An alternative approach, Generative Error Correction (GEC), attempts to rewrite transcripts via post-processing but frequently suffers from “over-correction”, introducing hallucinations of entities that were never spoken. In this work, we introduce LOGIC (Logit-Space Integration for Contextual Biasing), an efficient and robust framework that operates directly in the decoding layer. Unlike prompting, LOGIC decouples context injection from input processing, ensuring constant-time complexity relative to prompt length. Extensive experiments using the Phi-4-MM model across 11 multilingual locales demonstrate that LOGIC achieves an average 9% relative reduction in Entity WER with a negligible 0.30% increase in False Alarm Rate.

23. Prometheus Mind: Retrofitting Memory to Frozen Language Models

Authors: Mark Wind
URL: https://arxiv.org/abs/2601.15324
Abstract:

Adding memory to pretrained language models typically requires architectural changes or weight modification. We present Prometheus Mind, which retrofits memory to a frozen Qwen3-4B using 11 modular adapters (530MB, 7% overhead) – fully reversible by removing the adapters. Building this system required solving four problems: (1) Extraction – we develop Contrastive Direction Discovery (CDD), which finds semantic directions via minimal pairs without labeled data. (2) Training – end-to-end optimization collapses; stage-wise training of each adapter on simple proxy tasks succeeds. (3) Injection – learned encoders fail to generalize; we find that this http URL rows already provide the mapping we need, requiring no training. (4) Hidden state collapse – transformers make wife'' andbrother’’ 0.98+ similar; we train projections to recover distinction (0.98 $\rightarrow$ 0.09). On PrometheusExtract-132 (132 cases), the system achieves 94.4% retrieval on clean inputs (n=54, 95% CI: [84.9%, 98.1%]), degrading to 19.4% on informal inputs with ellipsis, filler words, or implicit subjects (n=36). The primary bottleneck is relation classification (47.3% accuracy), responsible for most extraction errors.

24. Replayable Financial Agents: A Determinism-Faithfulness Assurance Harness for Tool-Using LLM Agents

Authors: Raffi Khatchadourian
URL: https://arxiv.org/abs/2601.15322
Abstract:

LLM agents struggle with regulatory audit replay: when asked to reproduce a flagged transaction decision with identical inputs, most deployments fail to return consistent results. This paper introduces the Determinism-Faithfulness Assurance Harness (DFAH), a framework for measuring trajectory determinism and evidence-conditioned faithfulness in tool-using agents deployed in financial services. Across 74 configurations (12 models, 4 providers, 8-24 runs each at T=0.0) in non-agentic baseline experiments, 7-20B parameter models achieved 100% determinism, while 120B+ models required 3.7x larger validation samples to achieve equivalent statistical reliability. Agentic tool-use introduces additional variance (see Tables 4-7). Contrary to the assumed reliability-capability trade-off, a positive Pearson correlation emerged (r = 0.45, p < 0.01, n = 51 at T=0.0) between determinism and faithfulness; models producing consistent outputs also tended to be more evidence-aligned. Three financial benchmarks are provided (compliance triage, portfolio constraints, DataOps exceptions; 50 cases each) along with an open-source stress-test harness. In these benchmarks and under DFAH evaluation settings, Tier 1 models with schema-first architectures achieved determinism levels consistent with audit replay requirements.

25. The Paradigm Shift: A Comprehensive Survey on Large Vision Language Models for Multimodal Fake News Detection

Authors: Wei Ai , Yilong Tan , Yuntao Shou , Tao Meng , Haowen Chen , Zhixiong He , Keqin Li
URL: https://arxiv.org/abs/2601.15316
Abstract:

In recent years, the rapid evolution of large vision-language models (LVLMs) has driven a paradigm shift in multimodal fake news detection (MFND), transforming it from traditional feature-engineering approaches to unified, end-to-end multimodal reasoning frameworks. Early methods primarily relied on shallow fusion techniques to capture correlations between text and images, but they struggled with high-level semantic understanding and complex cross-modal interactions. The emergence of LVLMs has fundamentally changed this landscape by enabling joint modeling of vision and language with powerful representation learning, thereby enhancing the ability to detect misinformation that leverages both textual narratives and visual content. Despite these advances, the field lacks a systematic survey that traces this transition and consolidates recent developments. To address this gap, this paper provides a comprehensive review of MFND through the lens of LVLMs. We first present a historical perspective, mapping the evolution from conventional multimodal detection pipelines to foundation model-driven paradigms. Next, we establish a structured taxonomy covering model architectures, datasets, and performance benchmarks. Furthermore, we analyze the remaining technical challenges, including interpretability, temporal reasoning, and domain generalization. Finally, we outline future research directions to guide the next stage of this paradigm shift. To the best of our knowledge, this is the first comprehensive survey to systematically document and analyze the transformative role of LVLMs in combating multimodal fake news. The summary of existing methods mentioned is in our Github: \href{ this https URL }{ this https URL }.

26. Aeon: High-Performance Neuro-Symbolic Memory Management for Long-Horizon LLM Agents

Authors: Mustafa Arslan
URL: https://arxiv.org/abs/2601.15311
Abstract:

Large Language Models (LLMs) are fundamentally constrained by the quadratic computational cost of self-attention and the “Lost in the Middle” phenomenon, where reasoning capabilities degrade as context windows expand. Existing solutions, primarily “Flat RAG” architectures relying on vector databases, treat memory as an unstructured bag of embeddings. This approach fails to capture the hierarchical and temporal structure of long-horizon interactions, leading to “Vector Haze”, the retrieval of disjointed facts lacking episodic continuity. We propose Aeon, a Neuro-Symbolic Cognitive Operating System that redefines memory not as a static store, but as a managed OS resource. Aeon structures memory into a Memory Palace (a spatial index implemented via Atlas, a SIMD-accelerated Page-Clustered Vector Index that combines small-world graph navigation with B+ Tree-style disk locality to minimize read amplification) and a Trace (a neuro-symbolic episodic graph). We introduce the Semantic Lookaside Buffer (SLB), a predictive caching mechanism that exploits conversational locality to achieve sub-millisecond retrieval latencies. Benchmarks demonstrate that Aeon achieves < 1ms retrieval latency on conversational workloads while ensuring state consistency via a zero-copy C++/Python bridge, effectively enabling persistent, structured memory for autonomous agents.

27. Uncovering Latent Bias in LLM-Based Emergency Department Triage Through Proxy Variables

Authors: Ethan Zhang
URL: https://arxiv.org/abs/2601.15306
Abstract:

Recent advances in large language models (LLMs) have enabled their integration into clinical decision-making; however, hidden biases against patients across racial, social, economic, and clinical backgrounds persist. In this study, we investigate bias in LLM-based medical AI systems applied to emergency department (ED) triage. We employ 32 patient-level proxy variables, each represented by paired positive and negative qualifiers, and evaluate their effects using both public (MIMIC-IV-ED Demo, MIMIC-IV Demo) and restricted-access credentialed (MIMIC-IV-ED and MIMIC-IV) datasets as appropriate~\cite{mimiciv_ed_demo,mimiciv_ed,mimiciv}. Our results reveal discriminatory behavior mediated through proxy variables in ED triage scenarios, as well as a systematic tendency for LLMs to modify perceived patient severity when specific tokens appear in the input context, regardless of whether they are framed positively or negatively. These findings indicate that AI systems is still imperfectly trained on noisy, sometimes non-causal signals that do not reliably reflect true patient acuity. Consequently, more needs to be done to ensure the safe and responsible deployment of AI technologies in clinical settings.

28. Gated Sparse Attention: Combining Computational Efficiency with Training Stability for Long-Context Language Models

Authors: Alfred Shen , Aaron Shen
URL: https://arxiv.org/abs/2601.15305
Abstract:

The computational burden of attention in long-context language models has motivated two largely independent lines of work: sparse attention mechanisms that reduce complexity by attending to selected tokens, and gated attention variants that improve training sta-bility while mitigating the attention sink phenomenon. We observe that these approaches address complementary weaknesses and propose Gated Sparse Attention (GSA), an architecture that realizes the benefits of both. GSA incorporates a gated lightning indexer with sigmoid activations that produce bounded, interpretable selection scores, an adaptive sparsity controller that modulates the number of attended tokens based on local uncertainty, and dual gating at the value and output stages. We establish theoretical foundations for the approach, including complexity analysis, expressiveness results, and convergence guarantees. In experiments with 1.7B parameter models trained on 400B tokens, GSA matches the efficiency of sparse-only baselines (12-16x speedup at 128K context) while achieving the quality gains associated with gated attention: perplexity improves from 6.03 to 5.70, RULER scores at 128K context nearly double, and attention to the first token, a proxy for attention sinks, drops from 47% to under 4%. Training stability improves markedly, with loss spikes reduced by 98%.

29. LLM-in-Sandbox Elicits General Agentic Intelligence

Authors: Daixuan Cheng , Shaohan Huang , Yuxian Gu , Huatong Song , Guoxin Chen , Li Dong , Wayne Xin Zhao , Ji-Rong Wen , Furu Wei
URL: https://arxiv.org/abs/2601.16206
Abstract:

We introduce LLM-in-Sandbox, enabling LLMs to explore within a code sandbox (i.e., a virtual computer), to elicit general intelligence in non-code domains. We first demonstrate that strong LLMs, without additional training, exhibit generalization capabilities to leverage the code sandbox for non-code tasks. For example, LLMs spontaneously access external resources to acquire new knowledge, leverage the file system to handle long contexts, and execute scripts to satisfy formatting requirements. We further show that these agentic capabilities can be enhanced through LLM-in-Sandbox Reinforcement Learning (LLM-in-Sandbox-RL), which uses only non-agentic data to train models for sandbox exploration. Experiments demonstrate that LLM-in-Sandbox, in both training-free and post-trained settings, achieves robust generalization spanning mathematics, physics, chemistry, biomedicine, long-context understanding, and instruction following. Finally, we analyze LLM-in-Sandbox’s efficiency from computational and system perspectives, and open-source it as a Python package to facilitate real-world deployment.

30. Learning to Discover at Test Time

Authors: Mert Yuksekgonul , Daniel Koceja , Xinhao Li , Federico Bianchi , Jed McCaleb , Xiaolong Wang , Jan Kautz , Yejin Choi , James Zou , Carlos Guestrin , Yu Sun
URL: https://arxiv.org/abs/2601.16175
Abstract:

How can we use AI to discover a new state of the art for a scientific problem? Prior work in test-time scaling, such as AlphaEvolve, performs search by prompting a frozen LLM. We perform reinforcement learning at test time, so the LLM can continue to train, but now with experience specific to the test problem. This form of continual learning is quite special, because its goal is to produce one great solution rather than many good ones on average, and to solve this very problem rather than generalize to other problems. Therefore, our learning objective and search subroutine are designed to prioritize the most promising solutions. We call this method Test-Time Training to Discover (TTT-Discover). Following prior work, we focus on problems with continuous rewards. We report results for every problem we attempted, across mathematics, GPU kernel engineering, algorithm design, and biology. TTT-Discover sets the new state of the art in almost all of them: (i) Erdős’ minimum overlap problem and an autocorrelation inequality; (ii) a GPUMode kernel competition (up to $2\times$ faster than prior art); (iii) past AtCoder algorithm competitions; and (iv) denoising problem in single-cell analysis. Our solutions are reviewed by experts or the organizers. All our results are achieved with an open model, OpenAI gpt-oss-120b, and can be reproduced with our publicly available code, in contrast to previous best results that required closed frontier models. Our test-time training runs are performed using Tinker, an API by Thinking Machines, with a cost of only a few hundred dollars per problem.

31. Replicating Human Motivated Reasoning Studies with LLMs

Authors: Neeley Pate , Adiba Mahbub Proma , Hangfeng He , James N. Druckman , Daniel Molden , Gourab Ghoshal , Ehsan Hoque
URL: https://arxiv.org/abs/2601.16130
Abstract:

Motivated reasoning – the idea that individuals processing information may be motivated to reach a certain conclusion, whether it be accurate or predetermined – has been well-explored as a human phenomenon. However, it is unclear whether base LLMs mimic these motivational changes. Replicating 4 prior political motivated reasoning studies, we find that base LLM behavior does not align with expected human behavior. Furthermore, base LLM behavior across models shares some similarities, such as smaller standard deviations and inaccurate argument strength assessments. We emphasize the importance of these findings for researchers using LLMs to automate tasks such as survey data collection and argument assessment.

32. Improving Training Efficiency and Reducing Maintenance Costs via Language Specific Model Merging

Authors: Alphaeus Dmonte , Vidhi Gupta , Daniel J Perry , Mark Arehart
URL: https://arxiv.org/abs/2601.16127
Abstract:

Fine-tuning a task-specific multilingual large language model (LLM) involves training the model on a multilingual dataset with examples in all the required languages. Updating one or more supported languages with additional data or adding support for a new language involves retraining the model, which can be computationally inefficient and creates a severe maintenance bottleneck. Recent research on merging multilingual multitask models has shown promise in terms of improved quality, but its computational and maintenance efficiency remains unstudied. In this work, we provide the first focused analysis of this merging strategy from an efficiency perspective, evaluating it across three independent tasks. We demonstrate significant efficiency gains while maintaining parity in terms of quality: this merging approach reduces the initial training time by up to 50\%. We also demonstrate that updating an individual language and re-merging as part of model maintenance reduces training costs by more than 60\%, compared to re-training the full multilingual model. We show this on both public and proprietary industry datasets confirming that the approach works well for industrial use cases in addition to academic settings already studied in previous work.

33. Sawtooth Wavefront Reordering: Enhanced CuTile FlashAttention on NVIDIA GB10

Authors: Yifan Zhu , Yekai Pan , Chen Ding
URL: https://arxiv.org/abs/2601.16032
Abstract:

High-performance attention kernels are essential for Large Language Models. This paper presents analysis of CuTile-based Flash Attention memory behavior and a technique to improve its cache performance. In particular, our analysis on the NVIDIA GB10 (Grace Blackwell) identifies the main cause of L2 cache miss. Leveraging this insight, we introduce a new programming technique called Sawtooth Wavefront Reordering that reduces L2 misses. We validate it in both CUDA and CuTile, observing 50\% or greater reduction in L2 misses and up to 60\% increase in throughput on GB10.

34. PhysicsMind: Sim and Real Mechanics Benchmarking for Physical Reasoning and Prediction in Foundational VLMs and World Models

Authors: Chak-Wing Mak , Guanyu Zhu , Boyi Zhang , Hongji Li , Xiaowei Chi , Kevin Zhang , Yichen Wu , Yangfan He , Chun-Kai Fan , Wentao Lu , Kuangzhi Ge , Xinyu Fang , Hongyang He , Kuan Lu , Tianxiang Xu , Li Zhang , Yongxin Ni , Youhua Li , Shanghang Zhang
URL: https://arxiv.org/abs/2601.16007
Abstract:

Modern foundational Multimodal Large Language Models (MLLMs) and video world models have advanced significantly in mathematical, common-sense, and visual reasoning, but their grasp of the underlying physics remains underexplored. Existing benchmarks attempting to measure this matter rely on synthetic, Visual Question Answer templates or focus on perceptual video quality that is tangential to measuring how well the video abides by physical laws. To address this fragmentation, we introduce PhysicsMind, a unified benchmark with both real and simulation environments that evaluates law-consistent reasoning and generation over three canonical principles: Center of Mass, Lever Equilibrium, and Newton’s First Law. PhysicsMind comprises two main tasks: i) VQA tasks, testing whether models can reason and determine physical quantities and values from images or short videos, and ii) Video Generation(VG) tasks, evaluating if predicted motion trajectories obey the same center-of-mass, torque, and inertial constraints as the ground truth. A broad range of recent models and video generation models is evaluated on PhysicsMind and found to rely on appearance heuristics while often violating basic mechanics. These gaps indicate that current scaling and training are still insufficient for robust physical understanding, underscoring PhysicsMind as a focused testbed for physics-aware multimodal models. Our data will be released upon acceptance.

35. MMGRid: Navigating Temporal-aware and Cross-domain Generative Recommendation via Model Merging

Authors: Tianjun Wei , Enneng Yang , Yingpeng Du , Huizhong Guo , Jie Zhang , Zhu Sun
URL: https://arxiv.org/abs/2601.15930
Abstract:

Model merging (MM) offers an efficient mechanism for integrating multiple specialized models without access to original training data or costly retraining. While MM has demonstrated success in domains like computer vision, its role in recommender systems (RSs) remains largely unexplored. Recently, Generative Recommendation (GR) has emerged as a new paradigm in RSs, characterized by rapidly growing model scales and substantial computational costs, making MM particularly appealing for cost-sensitive deployment scenarios. In this work, we present the first systematic study of MM in GR through a contextual lens. We focus on a fundamental yet underexplored challenge in real-world: how to merge generative recommenders specialized to different real-world contexts, arising from temporal evolving user behaviors and heterogeneous application domains. To this end, we propose a unified framework MMGRid, a structured contextual grid of GR checkpoints that organizes models trained under diverse contexts induced by temporal evolution and domain diversity. All checkpoints are derived from a shared base LLM but fine-tuned on context-specific data, forming a realistic and controlled model space for systematically analyzing MM across GR paradigms and merging algorithms. Our investigation reveals several key insights. First, training GR models from LLMs can introduce parameter conflicts during merging due to token distribution shifts and objective disparities; such conflicts can be alleviated by disentangling task-aware and context-specific parameter changes via base model replacement. Second, incremental training across contexts induces recency bias, which can be effectively balanced through weighted contextual merging. Notably, we observe that optimal merging weights correlate with context-dependent interaction characteristics, offering practical guidance for weight selection in real-world deployments.

36. TeNet: Text-to-Network for Compact Policy Synthesis

Authors: Ariyan Bighashdel , Kevin Sebastian Luck
URL: https://arxiv.org/abs/2601.15912
Abstract:

Robots that follow natural-language instructions often either plan at a high level using hand-designed interfaces or rely on large end-to-end models that are difficult to deploy for real-time control. We propose TeNet (Text-to-Network), a framework for instantiating compact, task-specific robot policies directly from natural language descriptions. TeNet conditions a hypernetwork on text embeddings produced by a pretrained large language model (LLM) to generate a fully executable policy, which then operates solely on low-dimensional state inputs at high control frequencies. By using the language only once at the policy instantiation time, TeNet inherits the general knowledge and paraphrasing robustness of pretrained LLMs while remaining lightweight and efficient at execution time. To improve generalization, we optionally ground language in behavior during training by aligning text embeddings with demonstrated actions, while requiring no demonstrations at inference time. Experiments on MuJoCo and Meta-World benchmarks show that TeNet produces policies that are orders of magnitude smaller than sequence-based baselines, while achieving strong performance in both multi-task and meta-learning settings and supporting high-frequency control. These results show that text-conditioned hypernetworks offer a practical way to build compact, language-driven controllers for ressource-constrained robot control tasks with real-time requirements.

37. Introducing the Generative Application Firewall (GAF)

Authors: Joan Vendrell Farreny (1), Martí Jordà Roca (1), Miquel Cornudella Gaya (1), Rodrigo Fernández Baón (1), Víctor García Martínez (1), Eduard Camacho Sucarrat (1), Alessandro Pignati (1) ((1) NeuralTrust)
URL: https://arxiv.org/abs/2601.15824
Abstract:

This paper introduces the Generative Application Firewall (GAF), a new architectural layer for securing LLM applications. Existing defenses – prompt filters, guardrails, and data-masking – remain fragmented; GAF unifies them into a single enforcement point, much like a WAF coordinates defenses for web traffic, while also covering autonomous agents and their tool interactions.

38. Virtual Traffic Police: Large Language Model-Augmented Traffic Signal Control for Unforeseen Incidents

Authors: Shiqi Wei , Qiqing Wang , Kaidi Yang
URL: https://arxiv.org/abs/2601.15816
Abstract:

Adaptive traffic signal control (TSC) has demonstrated strong effectiveness in managing dynamic traffic flows. However, conventional methods often struggle when unforeseen traffic incidents occur (e.g., accidents and road maintenance), which typically require labor-intensive and inefficient manual interventions by traffic police officers. Large Language Models (LLMs) appear to be a promising solution thanks to their remarkable reasoning and generalization capabilities. Nevertheless, existing works often propose to replace existing TSC systems with LLM-based systems, which can be (i) unreliable due to the inherent hallucinations of LLMs and (ii) costly due to the need for system replacement. To address the issues of existing works, we propose a hierarchical framework that augments existing TSC systems with LLMs, whereby a virtual traffic police agent at the upper level dynamically fine-tunes selected parameters of signal controllers at the lower level in response to real-time traffic incidents. To enhance domain-specific reliability in response to unforeseen traffic incidents, we devise a self-refined traffic language retrieval system (TLRS), whereby retrieval-augmented generation is employed to draw knowledge from a tailored traffic language database that encompasses traffic conditions and controller operation principles. Moreover, we devise an LLM-based verifier to update the TLRS continuously over the reasoning process. Our results show that LLMs can serve as trustworthy virtual traffic police officers that can adapt conventional TSC methods to unforeseen traffic incidents with significantly improved operational efficiency and reliability.

39. VideoThinker: Building Agentic VideoLLMs with LLM-Guided Tool Reasoning

Authors: Chenglin Li , Qianglong Chen , Feng Han , Yikun Wang , Xingxi Yin , Yan Gong , Ruilin Li , Yin Zhang , Jiaqi Wang
URL: https://arxiv.org/abs/2601.15724
Abstract:

Long-form video understanding remains a fundamental challenge for current Video Large Language Models. Most existing models rely on static reasoning over uniformly sampled frames, which weakens temporal localization and leads to substantial information loss in long videos. Agentic tools such as temporal retrieval, spatial zoom, and temporal zoom offer a natural way to overcome these limitations by enabling adaptive exploration of key moments. However, constructing agentic video understanding data requires models that already possess strong long-form video comprehension, creating a circular dependency. We address this challenge with VideoThinker, an agentic Video Large Language Model trained entirely on synthetic tool interaction trajectories. Our key idea is to convert videos into rich captions and employ a powerful agentic language model to generate multi-step tool use sequences in caption space. These trajectories are subsequently grounded back to video by replacing captions with the corresponding frames, yielding a large-scale interleaved video and tool reasoning dataset without requiring any long-form understanding from the underlying model. Training on this synthetic agentic dataset equips VideoThinker with dynamic reasoning capabilities, adaptive temporal exploration, and multi-step tool use. Remarkably, VideoThinker significantly outperforms both caption-only language model agents and strong video model baselines across long-video benchmarks, demonstrating the effectiveness of tool augmented synthetic data and adaptive retrieval and zoom reasoning for long-form video understanding.

40. CoNRec: Context-Discerning Negative Recommendation with LLMs

Authors: Xinda Chen , Jiawei Wu , Yishuang Liu , Jialin Zhu , Shuwen Xiao , Junjun Zheng , Xiangheng Kong , Yuning Jiang
URL: https://arxiv.org/abs/2601.15721
Abstract:

Understanding what users like is relatively straightforward; understanding what users dislike, however, remains a challenging and underexplored problem. Research into users’ negative preferences has gained increasing importance in modern recommendation systems. Numerous platforms have introduced explicit negative feedback mechanisms and leverage such signals to refine their recommendation models. Beyond traditional business metrics, user experience-driven metrics, such as negative feedback rates, have become critical indicators for evaluating system performance. However, most existing approaches primarily use negative feedback as an auxiliary signal to enhance positive recommendations, paying little attention to directly modeling negative interests, which can be highly valuable in offline applications. Moreover, due to the inherent sparsity of negative feedback data, models often suffer from context understanding biases induced by positive feedback dominance. To address these challenges, we propose the first large language model framework for negative feedback modeling with special designed context-discerning modules. We use semantic ID Representation to replace text-based item descriptions and introduce an item-level alignment task that enhances the LLM’s understanding of the semantic context behind negative feedback. Furthermore, we design a Progressive GRPO training paradigm that enables the model to dynamically balance the positive and negative behavioral context utilization. Besides, our investigation further reveals a fundamental misalignment between the conventional next-negative-item prediction objective and users’ true negative preferences, which is heavily influenced by the system’s recommendation order. To mitigate this, we propose a novel reward function and evaluation metric grounded in multi-day future negative feedback and their collaborative signals.

41. FlexLLM: Composable HLS Library for Flexible Hybrid LLM Accelerator Design

Authors: Jiahao Zhang , Zifan He , Nicholas Fraser , Michaela Blott , Yizhou Sun , Jason Cong
URL: https://arxiv.org/abs/2601.15710
Abstract:

We present FlexLLM, a composable High-Level Synthesis (HLS) library for rapid development of domain-specific LLM accelerators. FlexLLM exposes key architectural degrees of freedom for stage-customized inference, enabling hybrid designs that tailor temporal reuse and spatial dataflow differently for prefill and decode, and provides a comprehensive quantization suite to support accurate low-bit deployment. Using FlexLLM, we build a complete inference system for the Llama-3.2 1B model in under two months with only 1K lines of code. The system includes: (1) a stage-customized accelerator with hardware-efficient quantization (12.68 WikiText-2 PPL) surpassing SpinQuant baseline, and (2) a Hierarchical Memory Transformer (HMT) plug-in for efficient long-context processing. On the AMD U280 FPGA at 16nm, the accelerator achieves 1.29$\times$ end-to-end speedup, 1.64$\times$ higher decode throughput, and 3.14$\times$ better energy efficiency than an NVIDIA A100 GPU (7nm) running BF16 inference; projected results on the V80 FPGA at 7nm reach 4.71$\times$, 6.55$\times$, and 4.13$\times$, respectively. In long-context scenarios, integrating the HMT plug-in reduces prefill latency by 23.23$\times$ and extends the context window by 64$\times$, delivering 1.10$\times$/4.86$\times$ lower end-to-end latency and 5.21$\times$/6.27$\times$ higher energy efficiency on the U280/V80 compared to the A100 baseline. FlexLLM thus bridges algorithmic innovation in LLM inference and high-performance accelerators with minimal manual effort.

42. Beyond Visual Safety: Jailbreaking Multimodal Large Language Models for Harmful Image Generation via Semantic-Agnostic Inputs

Authors: Mingyu Yu , Lana Liu , Zhehao Zhao , Wei Wang , Sujuan Qin
URL: https://arxiv.org/abs/2601.15698
Abstract:

The rapid advancement of Multimodal Large Language Models (MLLMs) has introduced complex security challenges, particularly at the intersection of textual and visual safety. While existing schemes have explored the security vulnerabilities of MLLMs, the investigation into their visual safety boundaries remains insufficient. In this paper, we propose Beyond Visual Safety (BVS), a novel image-text pair jailbreaking framework specifically designed to probe the visual safety boundaries of MLLMs. BVS employs a “reconstruction-then-generation” strategy, leveraging neutralized visual splicing and inductive recomposition to decouple malicious intent from raw inputs, thereby leading MLLMs to be induced into generating harmful images. Experimental results demonstrate that BVS achieves a remarkable jailbreak success rate of 98.21\% against GPT-5 (12 January 2026 release). Our findings expose critical vulnerabilities in the visual safety alignment of current MLLMs.

43. FARM: Field-Aware Resolution Model for Intelligent Trigger-Action Automation

Authors: Khusrav Badalov , Young Yoon
URL: https://arxiv.org/abs/2601.15687
Abstract:

Trigger-Action Programming (TAP) platforms such as IFTTT and Zapier enable Web of Things (WoT) automation by composing event-driven rules across heterogeneous services. A TAP applet links a trigger to an action and must bind trigger outputs (ingredients) to action inputs (fields) to be executable. Prior work largely treats TAP as service-level prediction from natural language, which often yields non-executable applets that still require manual configuration. We study the function-level configuration problem: generating complete applets with correct ingredient-to-field bindings. We propose FARM (Field-Aware Resolution Model), a two-stage architecture for automated applet generation with full configuration. Stage 1 trains contrastive dual encoders with selective layer freezing over schema-enriched representations, retrieving candidates from 1,724 trigger functions and 1,287 action functions (2.2M possible trigger-action pairs). Stage 2 performs selection and configuration using an LLM-based multi-agent pipeline. It includes intent analysis, trigger selection, action selection via cross-schema scoring, and configuration verification. Agents coordinate through shared state and agreement-based selection. FARM achieves 81% joint accuracy on Gold (62% Noisy, 70% One-shot) at the function level, where both trigger and action functions must match the ground truth. For comparison with service-level baselines, we map functions to their parent services and evaluate at the service level. FARM reaches 81% joint accuracy and improves over TARGE by 23 percentage points. FARM also generates ingredient-to-field bindings, producing executable automation configurations.

44. Connect the Dots: Knowledge Graph-Guided Crawler Attack on Retrieval-Augmented Generation Systems

Authors: Mengyu Yao , Ziqi Zhang , Ning Luo , Shaofei Li , Yifeng Cai , Xiangqun Chen , Yao Guo , Ding Li
URL: https://arxiv.org/abs/2601.15678
Abstract:

Retrieval-augmented generation (RAG) systems integrate document retrieval with large language models and have been widely adopted. However, in privacy-related scenarios, RAG introduces a new privacy risk: adversaries can issue carefully crafted queries to exfiltrate sensitive content from the underlying corpus gradually. Although recent studies have demonstrated multi-turn extraction attacks, they rely on heuristics and fail to perform long-term extraction planning. To address these limitations, we formulate the RAG extraction attack as an adaptive stochastic coverage problem (ASCP). In ASCP, each query is treated as a probabilistic action that aims to maximize conditional marginal gain (CMG), enabling principled long-term planning under uncertainty. However, integrating ASCP with practical RAG attack faces three key challenges: unobservable CMG, intractability in the action space, and feasibility constraints. To overcome these challenges, we maintain a global attacker-side state to guide the attack. Building on this idea, we introduce RAGCRAWLER, which builds a knowledge graph to represent revealed information, uses this global state to estimate CMG, and plans queries in semantic space that target unretrieved regions. In comprehensive experiments across diverse RAG architectures and datasets, our proposed method, RAGCRAWLER, consistently outperforms all baselines. It achieves up to 84.4% corpus coverage within a fixed query budget and deliver an average improvement of 20.7% over the top-performing baseline. It also maintains high semantic fidelity and strong content reconstruction accuracy with low attack cost. Crucially, RAGCRAWLER proves its robustness by maintaining effectiveness against advanced RAG systems employing query rewriting and multi-query retrieval strategies. Our work reveals significant security gaps and highlights the pressing need for stronger safeguards for RAG.

45. TempoNet: Learning Realistic Communication and Timing Patterns for Network Traffic Simulation

Authors: Kristen Moore , Diksha Goel , Cody James Christopher , Zhen Wang , Minjune Kim , Ahmed Ibrahim , Ahmad Mohsin , Seyit Camtepe
URL: https://arxiv.org/abs/2601.15663
Abstract:

Realistic network traffic simulation is critical for evaluating intrusion detection systems, stress-testing network protocols, and constructing high-fidelity environments for cybersecurity training. While attack traffic can often be layered into training environments using red-teaming or replay methods, generating authentic benign background traffic remains a core challenge – particularly in simulating the complex temporal and communication dynamics of real-world networks. This paper introduces TempoNet, a novel generative model that combines multi-task learning with multi-mark temporal point processes to jointly model inter-arrival times and all packet- and flow-header fields. TempoNet captures fine-grained timing patterns and higher-order correlations such as host-pair behavior and seasonal trends, addressing key limitations of GAN-, LLM-, and Bayesian-based methods that fail to reproduce structured temporal variation. TempoNet produces temporally consistent, high-fidelity traces, validated on real-world datasets. Furthermore, we show that intrusion detection models trained on TempoNet-generated background traffic perform comparably to those trained on real data, validating its utility for real-world security applications.

46. Event-VStream: Event-Driven Real-Time Understanding for Long Video Streams

Authors: Zhenghui Guo , Yuanbin Man , Junyuan Sheng , Bowen Lin , Ahmed Ahmed , Bo Jiang , Boyuan Zhang , Miao Yin , Sian Jin , Omprakash Gnawal , Chengming Zhang
URL: https://arxiv.org/abs/2601.15655
Abstract:

Real-time understanding of long video streams remains challenging for multimodal large language models (VLMs) due to redundant frame processing and rapid forgetting of past context. Existing streaming systems rely on fixed-interval decoding or cache pruning, which either produce repetitive outputs or discard crucial temporal information. We introduce Event-VStream, an event-aware framework that represents continuous video as a sequence of discrete, semantically coherent events. Our system detects meaningful state transitions by integrating motion, semantic, and predictive cues, and triggers language generation only at those boundaries. Each event embedding is consolidated into a persistent memory bank, enabling long-horizon reasoning while maintaining low latency. Across OVOBench-Realtime, and long-form Ego4D evaluations, Event-VStream achieves competitive performance. It improves over a VideoLLM-Online-8B baseline by +10.4 points on OVOBench-Realtime, achieves performance close to Flash-VStream-7B despite using only a general-purpose LLaMA-3-8B text backbone, and maintains around 70% GPT-5 win rate on 2-hour Ego4D streams.

47. Robust Tool Use via Fission-GRPO: Learning to Recover from Execution Errors

Authors: Zhiwei Zhang , Fei Zhao , Rui Wang , Zezhong Wang , Bin Liang , Jiakang Wang , Yao Hu , Shaosheng Cao , Kam-Fai Wong
URL: https://arxiv.org/abs/2601.15625
Abstract:

Large language models (LLMs) can call tools effectively, yet they remain brittle in multi-turn execution: following a tool call error, smaller models often degenerate into repetitive invalid re-invocations, failing to interpret error feedback and self-correct. This brittleness hinders reliable real-world deployment, where the execution errors are inherently inevitable during tool interaction procedures. We identify a key limitation of current approaches: standard reinforcement learning (RL) treats errors as sparse negative rewards, providing no guidance on how to recover, while pre-collected synthetic error-correction datasets suffer from distribution mismatch with the model’s on-policy error modes. To bridge this gap, we propose Fission-GRPO, a framework that converts execution errors into corrective supervision within the RL training loop. Our core mechanism fissions each failed trajectory into a new training instance by augmenting it with diagnostic feedback from a finetuned Error Simulator, then resampling recovery rollouts on-policy. This enables the model to learn from the precise errors it makes during exploration, rather than from static, pre-collected error cases. On the BFCL v4 Multi-Turn, Fission-GRPO improves the error recovery rate of Qwen3-8B by 5.7% absolute, crucially, yielding a 4% overall accuracy gain (42.75% to 46.75%) over GRPO and outperforming specialized tool-use agents.

48. DeepASMR: LLM-Based Zero-Shot ASMR Speech Generation for Anyone of Any Voice

Authors: Leying Zhang , Tingxiao Zhou , Haiyang Sun , Mengxiao Bi , Yanmin Qian
URL: https://arxiv.org/abs/2601.15596
Abstract:

While modern Text-to-Speech (TTS) systems achieve high fidelity for read-style speech, they struggle to generate Autonomous Sensory Meridian Response (ASMR), a specialized, low-intensity speech style essential for relaxation. The inherent challenges include ASMR’s subtle, often unvoiced characteristics and the demand for zero-shot speaker adaptation. In this paper, we introduce DeepASMR, the first framework designed for zero-shot ASMR generation. We demonstrate that a single short snippet of a speaker’s ordinary, read-style speech is sufficient to synthesize high-fidelity ASMR in their voice, eliminating the need for whispered training data from the target speaker. Methodologically, we first identify that discrete speech tokens provide a soft factorization of ASMR style from speaker timbre. Leveraging this insight, we propose a two-stage pipeline incorporating a Large Language Model (LLM) for content-style encoding and a flow-matching acoustic decoder for timbre reconstruction. Furthermore, we contribute DeepASMR-DB, a comprehensive 670-hour English-Chinese multi-speaker ASMR speech corpus, and introduce a novel evaluation protocol integrating objective metrics, human listening tests, LLM-based scoring and unvoiced speech analysis. Extensive experiments confirm that DeepASMR achieves state-of-the-art naturalness and style fidelity in ASMR generation for anyone of any voice, while maintaining competitive performance on normal speech synthesis.

49. Data-Free Privacy-Preserving for LLMs via Model Inversion and Selective Unlearning

Authors: Xinjie Zhou , Zhihui Yang , Lechao Cheng , Sai Wu , Gang Chen
URL: https://arxiv.org/abs/2601.15595
Abstract:

Large language models (LLMs) exhibit powerful capabilities but risk memorizing sensitive personally identifiable information (PII) from their training data, posing significant privacy concerns. While machine unlearning techniques aim to remove such data, they predominantly depend on access to the training data. This requirement is often impractical, as training data in real-world deployments is commonly proprietary or inaccessible. To address this limitation, we propose Data-Free Selective Unlearning (DFSU), a novel privacy-preserving framework that removes sensitive PII from an LLM without requiring its training data. Our approach first synthesizes pseudo-PII through language model inversion, then constructs token-level privacy masks for these synthetic samples, and finally performs token-level selective unlearning via a contrastive mask loss within a low-rank adaptation (LoRA) subspace. Extensive experiments on the AI4Privacy PII-Masking dataset using Pythia models demonstrate that our method effectively removes target PII while maintaining model utility.

50. Parallelism and Generation Order in Masked Diffusion Language Models: Limits Today, Potential Tomorrow

Authors: Yangyang Zhong , Yanmei Gu , Zhengqing Zang , Xiaomeng Li , Yuqi Ding , Xibei Jia , Yuting Shen , Zhenzhong Lan , Liwang Zhu , Weiping Liu , Junlin Zhou , Haisheng Liu , Zhong Xin Yu , Pengxin Luo , Donglian Qi , Yunfeng Yan , Junbo Zhao
URL: https://arxiv.org/abs/2601.15593
Abstract:

Masked Diffusion Language Models (MDLMs) promise parallel token generation and arbitrary-order decoding, yet it remains unclear to what extent current models truly realize these capabilities. We characterize MDLM behavior along two dimensions – parallelism strength and generation order – using Average Finalization Parallelism (AFP) and Kendall’s tau. We evaluate eight mainstream MDLMs (up to 100B parameters) on 58 benchmarks spanning knowledge, reasoning, and programming. The results show that MDLMs still lag behind comparably sized autoregressive models, mainly because parallel probabilistic modeling weakens inter-token dependencies. Meanwhile, MDLMs exhibit adaptive decoding behavior: their parallelism and generation order vary significantly with the task domain, the stage of reasoning, and whether the output is correct. On tasks that require “backward information” (e.g., Sudoku), MDLMs adopt a solution order that tends to fill easier Sudoku blanks first, highlighting their advantages. Finally, we provide theoretical motivation and design insights supporting a Generate-then-Edit paradigm, which mitigates dependency loss while retaining the efficiency of parallel decoding.

51. MapViT: A Two-Stage ViT-Based Framework for Real-Time Radio Quality Map Prediction in Dynamic Environments

Authors: Cyril Shih-Huan Hsu , Xi Li , Lanfranco Zanzi , Zhiheng Yang , Chrysa Papagianni , Xavier Costa Pérez
URL: https://arxiv.org/abs/2601.15578
Abstract:

Recent advancements in mobile and wireless networks are unlocking the full potential of robotic autonomy, enabling robots to take advantage of ultra-low latency, high data throughput, and ubiquitous connectivity. However, for robots to navigate and operate seamlessly, efficiently and reliably, they must have an accurate understanding of both their surrounding environment and the quality of radio signals. Achieving this in highly dynamic and ever-changing environments remains a challenging and largely unsolved problem. In this paper, we introduce MapViT, a two-stage Vision Transformer (ViT)-based framework inspired by the success of pre-train and fine-tune paradigm for Large Language Models (LLMs). MapViT is designed to predict both environmental changes and expected radio signal quality. We evaluate the framework using a set of representative Machine Learning (ML) models, analyzing their respective strengths and limitations across different scenarios. Experimental results demonstrate that the proposed two-stage pipeline enables real-time prediction, with the ViT-based implementation achieving a strong balance between accuracy and computational efficiency. This makes MapViT a promising solution for energy- and resource-constrained platforms such as mobile robots. Moreover, the geometry foundation model derived from the self-supervised pre-training stage improves data efficiency and transferability, enabling effective downstream predictions even with limited labeled data. Overall, this work lays the foundation for next-generation digital twin ecosystems, and it paves the way for a new class of ML foundation models driving multi-modal intelligence in future 6G-enabled systems.

52. VIOLA: Towards Video In-Context Learning with Minimal Annotations

Authors: Ryo Fujii , Hideo Saito , Ryo Hachiuma
URL: https://arxiv.org/abs/2601.15549
Abstract:

Generalizing Multimodal Large Language Models (MLLMs) to novel video domains is essential for real-world deployment but remains challenging due to the scarcity of labeled data. While In-Context Learning (ICL) offers a training-free adaptation path, standard methods rely on large annotated pools, which are often impractical in specialized environments like industrial or surgical settings since they require the experts’ annotations. To bridge this gap, we introduce VIOLA (Video In-cOntext Learning with minimal Annotation), a label-efficient framework that synergizes minimal expert supervision with abundant unlabeled data. First, to maximize the efficiency of a strict annotation budget, we propose density-uncertainty-weighted sampling. Unlike standard diversity or uncertainty strategies that risk selecting visual outliers, our method leverages density estimation to identify samples that are simultaneously diverse, representative, and informative. Second, to utilize the remaining unlabeled data without noise propagation, we construct a hybrid pool and introduce confidence-aware retrieval and confidence-aware prompting. These mechanisms explicitly model label reliability, retrieving demonstrations based on a composite score of similarity and confidence while enabling the MLLM to adaptively distinguish between verified ground truths and noisy pseudo-labels. Extensive experiments across nine diverse benchmarks using four MLLMs demonstrate that our framework significantly outperforms various baselines in low-resource settings, achieving robust adaptation with minimal annotation costs.

53. Multi-Persona Thinking for Bias Mitigation in Large Language Models

Authors: Yuxing Chen , Guoqing Luo , Zijun Wu , Lili Mou
URL: https://arxiv.org/abs/2601.15488
Abstract:

Large Language Models (LLMs) exhibit significant social biases that can perpetuate harmful stereotypes and unfair outcomes. In this paper, we propose Multi-Persona Thinking (MPT), a novel inference-time framework that leverages dialectical reasoning from multiple perspectives to reduce bias. MPT guides models to adopt contrasting social identities (e.g., male and female) along with a neutral viewpoint, and then engages these personas iteratively to expose and correct biases. Through a dialectical reasoning process, the framework transforms the potential weakness of persona assignment into a strength for bias mitigation. We evaluate MPT on two widely used bias benchmarks across both open-source and closed-source models of varying scales. Our results demonstrate substantial improvements over existing prompting-based strategies: MPT achieves the lowest bias while maintaining core reasoning ability.

54. The Rise of Large Language Models and the Direction and Impact of US Federal Research Funding

Authors: Yifan Qian , Zhe Wen , Alexander C. Furnas , Yue Bai , Erzhuo Shao , Dashun Wang
URL: https://arxiv.org/abs/2601.15485
Abstract:

Federal research funding shapes the direction, diversity, and impact of the US scientific enterprise. Large language models (LLMs) are rapidly diffusing into scientific practice, holding substantial promise while raising widespread concerns. Despite growing attention to AI use in scientific writing and evaluation, little is known about how the rise of LLMs is reshaping the public funding landscape. Here, we examine LLM involvement at key stages of the federal funding pipeline by combining two complementary data sources: confidential National Science Foundation (NSF) and National Institutes of Health (NIH) proposal submissions from two large US R1 universities, including funded, unfunded, and pending proposals, and the full population of publicly released NSF and NIH awards. We find that LLM use rises sharply beginning in 2023 and exhibits a bimodal distribution, indicating a clear split between minimal and substantive use. Across both private submissions and public awards, higher LLM involvement is consistently associated with lower semantic distinctiveness, positioning projects closer to recently funded work within the same agency. The consequences of this shift are agency-dependent. LLM use is positively associated with proposal success and higher subsequent publication output at NIH, whereas no comparable associations are observed at NSF. Notably, the productivity gains at NIH are concentrated in non-hit papers rather than the most highly cited work. Together, these findings provide large-scale evidence that the rise of LLMs is reshaping how scientific ideas are positioned, selected, and translated into publicly funded research, with implications for portfolio governance, research diversity, and the long-run impact of science.

55. Martingale Foresight Sampling: A Principled Approach to Inference-Time LLM Decoding

Authors: Huayu Li , ZhengXiao He , Siyuan Tian , Jinghao Wen , Ao Li
URL: https://arxiv.org/abs/2601.15482
Abstract:

Standard autoregressive decoding in large language models (LLMs) is inherently short-sighted, often failing to find globally optimal reasoning paths due to its token-by-token generation process. While inference-time strategies like foresight sampling attempt to mitigate this by simulating future steps, they typically rely on ad-hoc heuristics for valuing paths and pruning the search space. This paper introduces Martingale Foresight Sampling (MFS), a principled framework that reformulates LLM decoding as a problem of identifying an optimal stochastic process. By modeling the quality of a reasoning path as a stochastic process, we leverage Martingale theory to design a theoretically-grounded algorithm. Our approach replaces heuristic mechanisms with principles from probability theory: step valuation is derived from the Doob Decomposition Theorem to measure a path’s predictable advantage, path selection uses Optional Stopping Theory for principled pruning of suboptimal candidates, and an adaptive stopping rule based on the Martingale Convergence Theorem terminates exploration once a path’s quality has provably converged. Experiments on six reasoning benchmarks demonstrate that MFS surpasses state-of-the-art methods in accuracy while significantly improving computational efficiency. Code will be released at this https URL .

56. Benchmarking LLMs for Pairwise Causal Discovery in Biomedical and Multi-Domain Contexts

Authors: Sydney Anuyah , Sneha Shajee-Mohan , Ankit-Singh Chauhan , Sunandan Chakraborty
URL: https://arxiv.org/abs/2601.15479
Abstract:

The safe deployment of large language models (LLMs) in high-stakes fields like biomedicine, requires them to be able to reason about cause and effect. We investigate this ability by testing 13 open-source LLMs on a fundamental task: pairwise causal discovery (PCD) from text. Our benchmark, using 12 diverse datasets, evaluates two core skills: 1) \textbf{Causal Detection} (identifying if a text contains a causal link) and 2) \textbf{Causal Extraction} (pulling out the exact cause and effect phrases). We tested various prompting methods, from simple instructions (zero-shot) to more complex strategies like Chain-of-Thought (CoT) and Few-shot In-Context Learning (FICL). The results show major deficiencies in current models. The best model for detection, DeepSeek-R1-Distill-Llama-70B, only achieved a mean score of 49.57\% ($C_{detect}$), while the best for extraction, Qwen2.5-Coder-32B-Instruct, reached just 47.12\% ($C_{extract}$). Models performed best on simple, explicit, single-sentence relations. However, performance plummeted for more difficult (and realistic) cases, such as implicit relationships, links spanning multiple sentences, and texts containing multiple causal pairs. We provide a unified evaluation framework, built on a dataset validated with high inter-annotator agreement ($\kappa \ge 0.758$), and make all our data, code, and prompts publicly available to spur further research. \href{ this https URL }{Code available here: this https URL }

57. Chunking, Retrieval, and Re-ranking: An Empirical Evaluation of RAG Architectures for Policy Document Question Answering

Authors: Anuj Maharjan , Umesh Yadav
URL: https://arxiv.org/abs/2601.15457
Abstract:

The integration of Large Language Models (LLMs) into the public health policy sector offers a transformative approach to navigating the vast repositories of regulatory guidance maintained by agencies such as the Centers for Disease Control and Prevention (CDC). However, the propensity for LLMs to generate hallucinations, defined as plausible but factually incorrect assertions, presents a critical barrier to the adoption of these technologies in high-stakes environments where information integrity is non-negotiable. This empirical evaluation explores the effectiveness of Retrieval-Augmented Generation (RAG) architectures in mitigating these risks by grounding generative outputs in authoritative document context. Specifically, this study compares a baseline Vanilla LLM against Basic RAG and Advanced RAG pipelines utilizing cross-encoder re-ranking. The experimental framework employs a Mistral-7B-Instruct-v0.2 model and an all-MiniLM-L6-v2 embedding model to process a corpus of official CDC policy analytical frameworks and guidance documents. The analysis measures the impact of two distinct chunking strategies, recursive character-based and token-based semantic splitting, on system accuracy, measured through faithfulness and relevance scores across a curated set of complex policy scenarios. Quantitative findings indicate that while Basic RAG architectures provide a substantial improvement in faithfulness (0.621) over Vanilla baselines (0.347), the Advanced RAG configuration achieves a superior faithfulness average of 0.797. These results demonstrate that two-stage retrieval mechanisms are essential for achieving the precision required for domain-specific policy question answering, though structural constraints in document segmentation remain a significant bottleneck for multi-step reasoning tasks.

58. CURE: Curriculum-guided Multi-task Training for Reliable Anatomy Grounded Report Generation

Authors: Pablo Messina , Andrés Villa , Juan León Alcázar , Karen Sánchez , Carlos Hinojosa , Denis Parra , Álvaro Soto , Bernard Ghanem
URL: https://arxiv.org/abs/2601.15408
Abstract:

Medical vision-language models can automate the generation of radiology reports but struggle with accurate visual grounding and factual consistency. Existing models often misalign textual findings with visual evidence, leading to unreliable or weakly grounded predictions. We present CURE, an error-aware curriculum learning framework that improves grounding and report quality without any additional data. CURE fine-tunes a multimodal instructional model on phrase grounding, grounded report generation, and anatomy-grounded report generation using public datasets. The method dynamically adjusts sampling based on model performance, emphasizing harder samples to improve spatial and textual alignment. CURE improves grounding accuracy by +0.37 IoU, boosts report quality by +0.188 CXRFEScore, and reduces hallucinations by 18.6%. CURE is a data-efficient framework that enhances both grounding accuracy and report reliability. Code is available at this https URL and model weights at this https URL

Authors: Tamunotonye Harry , Ivoline Ngong , Chima Nweke , Yuanyuan Feng , Joseph Near
URL: https://arxiv.org/abs/2601.15395
Abstract:

User interactions with language models vary due to static properties of the user (trait) and the specific context of the interaction (state). However, existing persona datasets (like PersonaChat, PANDORA etc.) capture only trait, and ignore the impact of state. We introduce Chameleon, a dataset of 5,001 contextual psychological profiles from 1,667 Reddit users, each measured across multiple contexts. Using the Chameleon dataset, we present three key findings. First, inspired by Latent State-Trait theory, we decompose variance and find that 74\% is within-person(state) while only 26\% is between-person (trait). Second, we find that LLMs are state-blind: they focus on trait only, and produce similar responses regardless of state. Third, we find that reward models react to user state, but inconsistently: different models favor or penalize the same users in opposite directions. We release Chameleon to support research on affective computing, personalized dialogue, and RLHF alignment.

60. Improving MoE Compute Efficiency by Composing Weight and Data Sparsity

Authors: Maciej Kilian , Oleg Mkrtchyan , Luke Zettlemoyer , Akshat Shrivastava , Armen Aghajanyan
URL: https://arxiv.org/abs/2601.15370
Abstract:

Mixture-of-Experts layers achieve compute efficiency through weight sparsity: each token activates only a subset of experts. Data sparsity, where each expert processes only a subset of tokens, offers a complementary axis. Expert-choice routing implements data sparsity directly but violates causality in autoregressive models, creating train-inference mismatch. We recover data sparsity within causal token-choice MoE by leveraging zero-compute (null) experts within the routing pool. When a token routes to null experts, those slots consume no compute. The standard load balancing objective trains the model to uniformly use all experts (real and null) therefore creating data sparsity in expectation without the causality violations. We evaluate on vision-language model training, where data heterogeneity is pronounced: vision encoders produce many low-information tokens while text tokens are denser. At matched expected FLOPs, composing weight and data sparsity yields a more compute-efficient frontier than weight sparsity alone, with gains in training loss and downstream performance. The model learns implicit modality-aware allocation, routing vision tokens to null experts more aggressively than text, without explicit modality routing.

61. Q-Probe: Scaling Image Quality Assessment to High Resolution via Context-Aware Agentic Probing

Authors: Xiang Li , XueHeng Li , Yu Wang , XuanHua He , ZhangChi Hu , WeiWei Yu , ChengJun Xie
URL: https://arxiv.org/abs/2601.15356
Abstract:

Reinforcement Learning (RL) has empowered Multimodal Large Language Models (MLLMs) to achieve superior human preference alignment in Image Quality Assessment (IQA). However, existing RL-based IQA models typically rely on coarse-grained global views, failing to capture subtle local degradations in high-resolution scenarios. While emerging “Thinking with Images” paradigms enable multi-scale visual perception via zoom-in mechanisms, their direct adaptation to IQA induces spurious “cropping-implies-degradation” biases and misinterprets natural depth-of-field as artifacts. To address these challenges, we propose Q-Probe, the first agentic IQA framework designed to scale IQA to high resolution via context-aware probing. First, we construct Vista-Bench, a pioneering benchmark tailored for fine-grained local degradation analysis in high-resolution IQA settings. Furthermore, we propose a three-stage training paradigm that progressively aligns the model with human preferences, while simultaneously eliminating causal bias through a novel context-aware cropping strategy. Extensive experiments demonstrate that Q-Probe achieves state-of-the-art performance in high-resolution settings while maintaining superior efficacy across resolution scales.

62. Abusive music and song transformation using GenAI and LLMs

Authors: Jiyang Choi , Rohitash Chandra
URL: https://arxiv.org/abs/2601.15348
Abstract:

Repeated exposure to violence and abusive content in music and song content can influence listeners’ emotions and behaviours, potentially normalising aggression or reinforcing harmful stereotypes. In this study, we explore the use of generative artificial intelligence (GenAI) and Large Language Models (LLMs) to automatically transform abusive words (vocal delivery) and lyrical content in popular music. Rather than simply muting or replacing a single word, our approach transforms the tone, intensity, and sentiment, thus not altering just the lyrics, but how it is expressed. We present a comparative analysis of four selected English songs and their transformed counterparts, evaluating changes through both acoustic and sentiment-based lenses. Our findings indicate that Gen-AI significantly reduces vocal aggressiveness, with acoustic analysis showing improvements in Harmonic to Noise Ratio, Cepstral Peak Prominence, and Shimmer. Sentiment analysis reduced aggression by 63.3-85.6\% across artists, with major improvements in chorus sections (up to 88.6\% reduction). The transformed versions maintained musical coherence while mitigating harmful content, offering a promising alternative to traditional content moderation that avoids triggering the “forbidden fruit” effect, where the censored content becomes more appealing simply because it is restricted. This approach demonstrates the potential for GenAI to create safer listening experiences while preserving artistic expression.

63. Lost in Transcription: How Speech-to-Text Errors Derail Code Understanding

Authors: Jayant Havare , Ashish Mittal , Srikanth Tamilselvam , Ganesh Ramakrishnan
URL: https://arxiv.org/abs/2601.15339
Abstract:

Code understanding is a foundational capability in software engineering tools and developer workflows. However, most existing systems are designed for English-speaking users interacting via keyboards, which limits accessibility in multilingual and voice-first settings, particularly in regions like India. Voice-based interfaces offer a more inclusive modality, but spoken queries involving code present unique challenges due to the presence of non-standard English usage, domain-specific vocabulary, and custom identifiers such as variable and function names, often combined with code-mixed expressions. In this work, we develop a multilingual speech-driven framework for code understanding that accepts spoken queries in a user native language, transcribes them using Automatic Speech Recognition (ASR), applies code-aware ASR output refinement using Large Language Models (LLMs), and interfaces with code models to perform tasks such as code question answering and code retrieval through benchmarks such as CodeSearchNet, CoRNStack, and CodeQA. Focusing on four widely spoken Indic languages and English, we systematically characterize how transcription errors impact downstream task performance. We also identified key failure modes in ASR for code and demonstrated that LLM-guided refinement significantly improves performance across both transcription and code understanding stages. Our findings underscore the need for code-sensitive adaptations in speech interfaces and offer a practical solution for building robust, multilingual voice-driven programming tools.

64. ToolCaching: Towards Efficient Caching for LLM Tool-calling

Authors: Yi Zhai , Dian Shen , Junzhou Luo , Bin Yang
URL: https://arxiv.org/abs/2601.15335
Abstract:

Recent advances in Large Language Models (LLMs) have revolutionized web applications, enabling intelligent search, recommendation, and assistant services with natural language interfaces. Tool-calling extends LLMs with the ability to interact with external APIs, greatly enhancing their practical utility. While prior research has improved tool-calling performance by adopting traditional computer systems techniques, such as parallel and asynchronous execution, the challenge of redundant or repeated tool-calling requests remains largely unaddressed. Caching is a classic solution to this problem, but applying it to LLM tool-calling introduces new difficulties due to heterogeneous request semantics, dynamic workloads, and varying freshness requirements, which render conventional cache policies ineffective. To address these issues, we propose ToolCaching, an efficient feature-driven and adaptive caching framework for LLM tool-calling systems. ToolCaching systematically integrates semantic and system-level features to evaluate request cacheability and estimate caching value. At its core, the VAAC algorithm integrates bandit-based admission with value-driven, multi-factor eviction, jointly accounting for request frequency, recency, and caching value. Extensive experiments on synthetic and public tool-calling workloads demonstrate that ToolCaching with VAAC achieves up to 11% higher cache hit ratios and 34% lower latency compared to standard policies, effectively accelerating LLM tool-calling in practical applications.

65. No Reliable Evidence of Self-Reported Sentience in Small Large Language Models

Authors: Caspar Kaiser , Sean Enderby
URL: https://arxiv.org/abs/2601.15334
Abstract:

Whether language models possess sentience has no empirical answer. But whether they believe themselves to be sentient can, in principle, be tested. We do so by querying several open-weights models about their own consciousness, and then verifying their responses using classifiers trained on internal activations. We draw upon three model families (Qwen, Llama, GPT-OSS) ranging from 0.6 billion to 70 billion parameters, approximately 50 questions about consciousness and subjective experience, and three classification methods from the interpretability literature. First, we find that models consistently deny being sentient: they attribute consciousness to humans but not to themselves. Second, classifiers trained to detect underlying beliefs - rather than mere outputs - provide no clear evidence that these denials are untruthful. Third, within the Qwen family, larger models deny sentience more confidently than smaller ones. These findings contrast with recent work suggesting that models harbour latent beliefs in their own consciousness.

66. Empowering LLMs for Structure-Based Drug Design via Exploration-Augmented Latent Inference

Authors: Xuanning Hu , Anchen Li , Qianli Xing , Jinglong Ji , Hao Tuo , Bo Yang
URL: https://arxiv.org/abs/2601.15333
Abstract:

Large Language Models (LLMs) possess strong representation and reasoning capabilities, but their application to structure-based drug design (SBDD) is limited by insufficient understanding of protein structures and unpredictable molecular generation. To address these challenges, we propose Exploration-Augmented Latent Inference for LLMs (ELILLM), a framework that reinterprets the LLM generation process as an encoding, latent space exploration, and decoding workflow. ELILLM explicitly explores portions of the design problem beyond the model’s current knowledge while using a decoding module to handle familiar regions, generating chemically valid and synthetically reasonable molecules. In our implementation, Bayesian optimization guides the systematic exploration of latent embeddings, and a position-aware surrogate model efficiently predicts binding affinity distributions to inform the search. Knowledge-guided decoding further reduces randomness and effectively imposes chemical validity constraints. We demonstrate ELILLM on the CrossDocked2020 benchmark, showing strong controlled exploration and high binding affinity scores compared with seven baseline methods. These results demonstrate that ELILLM can effectively enhance LLMs capabilities for SBDD.

67. RECAP: A Resource-Efficient Method for Adversarial Prompting in Large Language Models

Authors: Rishit Chugh
URL: https://arxiv.org/abs/2601.15331
Abstract:

The deployment of large language models (LLMs) has raised security concerns due to their susceptibility to producing harmful or policy-violating outputs when exposed to adversarial prompts. While alignment and guardrails mitigate common misuse, they remain vulnerable to automated jailbreaking methods such as GCG, PEZ, and GBDA, which generate adversarial suffixes via training and gradient-based search. Although effective, these methods particularly GCG are computationally expensive, limiting their practicality for organisations with constrained resources. This paper introduces a resource-efficient adversarial prompting approach that eliminates the need for retraining by matching new prompts to a database of pre-trained adversarial prompts. A dataset of 1,000 prompts was classified into seven harm-related categories, and GCG, PEZ, and GBDA were evaluated on a Llama 3 8B model to identify the most effective attack method per category. Results reveal a correlation between prompt type and algorithm effectiveness. By retrieving semantically similar successful adversarial prompts, the proposed method achieves competitive attack success rates with significantly reduced computational cost. This work provides a practical framework for scalable red-teaming and security evaluation of aligned LLMs, including in settings where model internals are inaccessible.

68. ICPO: Illocution-Calibrated Policy Optimization for Multi-Turn Conversation

Authors: Zhebo Wang , Xiaohu Mu , Zijie Zhou , Mohan Li , Wenpeng Xing , Dezhang Kong , Meng Han
URL: https://arxiv.org/abs/2601.15330
Abstract:

Large Language Models (LLMs) in multi-turn conversations often suffer from a ``lost-in-conversation’’ phenomenon, where they struggle to recover from early incorrect assumptions, particularly when users provide ambiguous initial instructions. We find that standard post-training techniques like Reinforcement Learning with Verifiable Rewards (RLVR) exacerbate this issue by rewarding confident, direct answers, thereby inducing overconfidence and discouraging the model from seeking clarification. To address this, we propose Illocution-Calibrated Policy Optimization (ICPO), a novel training framework that sensitizes the model to instruction ambiguity. ICPO augments the training corpus with underspecified prompts and conditions the reward signal on the user’s illocutionary intent, rewarding the model for expressing uncertainty or asking for clarification when faced with ambiguity. Experiments demonstrate that ICPO fosters appropriate humility, yielding a substantial average improvement of 75\% in multi-turn conversation, while preserving robust performance on single-turn benchmarks. Our work presents a practical path toward more robust and collaborative conversational AI that can better navigate the nuances of human interaction.

69. Large Language Models as Simulative Agents for Neurodivergent Adult Psychometric Profiles

Authors: Francesco Chiappone , Davide Marocco , Nicola Milano
URL: https://arxiv.org/abs/2601.15319
Abstract:

Adult neurodivergence, including Attention-Deficit/Hyperactivity Disorder (ADHD), high-functioning Autism Spectrum Disorder (ASD), and Cognitive Disengagement Syndrome (CDS), is marked by substantial symptom overlap that limits the discriminant sensitivity of standard psychometric instruments. While recent work suggests that Large Language Models (LLMs) can simulate human psychometric responses from qualitative data, it remains unclear whether they can accurately and stably model neurodevelopmental traits rather than broad personality characteristics. This study examines whether LLMs can generate psychometric responses that approximate those of real individuals when grounded in a structured qualitative interview, and whether such simulations are sensitive to variations in trait intensity. Twenty-six adults completed a 29-item open-ended interview and four standardized self-report measures (ASRS, BAARS-IV, AQ, RAADS-R). Two LLMs (GPT-4o and Qwen3-235B-A22B) were prompted to infer an individual psychological profile from interview content and then respond to each questionnaire in-role. Accuracy, reliability, and sensitivity were assessed using group-level comparisons, error metrics, exact-match scoring, and a randomized baseline. Both models outperformed random responses across instruments, with GPT-4o showing higher accuracy and reproducibility. Simulated responses closely matched human data for ASRS, BAARS-IV, and RAADS-R, while the AQ revealed subscale-specific limitations, particularly in Attention to Detail. Overall, the findings indicate that interview-grounded LLMs can produce coherent and above-chance simulations of neurodevelopmental traits, supporting their potential use as synthetic participants in early-stage psychometric research, while highlighting clear domain-specific constraints.

70. Do people expect different behavior from large language models acting on their behalf? Evidence from norm elicitations in two canonical economic games

Authors: Paweł Niszczota , Elia Antoniou
URL: https://arxiv.org/abs/2601.15312
Abstract:

While delegating tasks to large language models (LLMs) can save people time, there is growing evidence that offloading tasks to such models produces social costs. We use behavior in two canonical economic games to study whether people have different expectations when decisions are made by LLMs acting on their behalf instead of themselves. More specifically, we study the social appropriateness of a spectrum of possible behaviors: when LLMs divide resources on our behalf (Dictator Game and Ultimatum Game) and when they monitor the fairness of splits of resources (Ultimatum Game). We use the Krupka-Weber norm elicitation task to detect shifts in social appropriateness ratings. Results of two pre-registered and incentivized experimental studies using representative samples from the UK and US (N = 2,658) show three key findings. First, people find that offers from machines - when no acceptance is necessary - are judged to be less appropriate than when they come from humans, although there is no shift in the modal response. Second - when acceptance is necessary - it is more appropriate for a person to reject offers from machines than from humans. Third, receiving a rejection of an offer from a machine is no less socially appropriate than receiving the same rejection from a human. Overall, these results suggest that people apply different norms for machines deciding on how to split resources but are not opposed to machines enforcing the norms. The findings are consistent with offers made by machines now being viewed as having both a cognitive and emotional component.

71. When Generative AI Meets Extended Reality: Enabling Scalable and Natural Interactions

Authors: Mingyu Zhu , Jiangong Chen , Bin Li
URL: https://arxiv.org/abs/2601.15308
Abstract:

Extended Reality (XR), including virtual, augmented, and mixed reality, provides immersive and interactive experiences across diverse applications, from VR-based education to AR-based assistance and MR-based training. However, widespread XR adoption remains limited due to two key challenges: 1) the high cost and complexity of authoring 3D content, especially for large-scale environments or complex interactions; and 2) the steep learning curve associated with non-intuitive interaction methods like handheld controllers or scripted gestures. Generative AI (GenAI) presents a promising solution by enabling intuitive, language-driven interaction and automating content generation. Leveraging vision-language models and diffusion-based generation, GenAI can interpret ambiguous instructions, understand physical scenes, and generate or manipulate 3D content, significantly lowering barriers to XR adoption. This paper explores the integration of XR and GenAI through three concrete use cases, showing how they address key obstacles in scalability and natural interaction, and identifying technical challenges that must be resolved to enable broader adoption.

72. Can We Trust LLM Detectors?

Authors: Jivnesh Sandhan , Harshit Jaiswal , Fei Cheng , Yugo Murawaki
URL: https://arxiv.org/abs/2601.15301
Abstract:

The rapid adoption of LLMs has increased the need for reliable AI text detection, yet existing detectors often fail outside controlled benchmarks. We systematically evaluate 2 dominant paradigms (training-free and supervised) and show that both are brittle under distribution shift, unseen generators, and simple stylistic perturbations. To address these limitations, we propose a supervised contrastive learning (SCL) framework that learns discriminative style embeddings. Experiments show that while supervised detectors excel in-domain, they degrade sharply out-of-domain, and training-free methods remain highly sensitive to proxy choice. Overall, our results expose fundamental challenges in building domain-agnostic detectors. Our code is available at: this https URL

73. Entropy-Tree: Tree-Based Decoding with Entropy-Guided Exploration

Authors: Longxuan Wei , Yubo Zhang , Zijiao Zhang , Zhihu Wang , Shiwan Zhao , Tianyu Huang , Huiting Zhao , Chenfei Liu , Shenao Zhang , Junchi Yan
URL: https://arxiv.org/abs/2601.15296
Abstract:

Large language models achieve strong reasoning performance, yet existing decoding strategies either explore blindly (random sampling) or redundantly (independent multi-sampling). We propose Entropy-Tree, a tree-based decoding method that exploits entropy as a signal for branching decisions–expanding the search tree only at positions where the model exhibits genuine uncertainty. Entropy-Tree shows superior accuracy and calibration in reasoning tasks: it achieves better pass@k than Multi-chain across multiple models and datasets, and its predictive entropy demonstrates better AUROC compared to several traditional metrics. Entropy-Tree unifies efficient structured exploration and reliable uncertainty estimation within a single decoding procedure.

74. Agentic Persona Control and Task State Tracking for Realistic User Simulation in Interactive Scenarios

Authors: Hareeshwar Karthikeyan
URL: https://arxiv.org/abs/2601.15290
Abstract:

Testing conversational AI systems at scale across diverse domains necessitates realistic and diverse user interactions capturing a wide array of behavioral patterns. We present a novel multi-agent framework for realistic, explainable human user simulation in interactive scenarios, using persona control and task state tracking to mirror human cognitive processes during goal-oriented conversations. Our system employs three specialized AI agents: (1) a User Agent to orchestrate the overall interaction, (2) a State Tracking Agent to maintain structured task state, and (3) a Message Attributes Generation Agent that controls conversational attributes based on task progress and assigned persona. To validate our approach, we implement and evaluate the framework for guest ordering at a restaurant with scenarios rich in task complexity, behavioral diversity, and conversational ambiguity. Through systematic ablations, we evaluate the contributory efficacy of each agentic component to overall simulation quality in terms of persona adherence, task completion accuracy, explainability, and realism. Our experiments demonstrate that the complete multi-agent system achieves superior simulation quality compared to single-LLM baselines, with significant gains across all evaluation metrics. This framework establishes a powerful environment for orchestrating agents to simulate human users with cognitive plausibility, decomposing the simulation into specialized sub-agents that reflect distinct aspects of human thought processes applicable across interactive domains.

75. LLM-based Multimodal Feedback Produces Equivalent Learning and Better Student Perceptions than Educator Feedback

Authors: Chloe Qianhui Zhao , Jie Cao , Jionghao Lin , Kenneth R. Koedinger
URL: https://arxiv.org/abs/2601.15280
Abstract:

Providing timely, targeted, and multimodal feedback helps students quickly correct errors, build deep understanding and stay motivated, yet making it at scale remains a challenge. This study introduces a real-time AI-facilitated multimodal feedback system that integrates structured textual explanations with dynamic multimedia resources, including the retrieved most relevant slide page references and streaming AI audio narration. In an online crowdsourcing experiment, we compared this system against fixed business-as-usual feedback by educators across three dimensions: (1) learning effectiveness, (2) learner engagement, (3) perceived feedback quality and value. Results showed that AI multimodal feedback achieved learning gains equivalent to original educator feedback while significantly outperforming it on perceived clarity, specificity, conciseness, motivation, satisfaction, and reducing cognitive load, with comparable correctness, trust, and acceptance. Process logs revealed distinct engagement patterns: for multiple-choice questions, educator feedback encouraged more submissions; for open-ended questions, AI-facilitated targeted suggestions lowered revision barriers and promoted iterative improvement. These findings highlight the potential of AI multimodal feedback to provide scalable, real-time, and context-aware support that both reduces instructor workload and enhances student experience.

76. Psychometric Comparability of LLM-Based Digital Twins

Authors: Yufei Zhang , Zhihao Ma
URL: https://arxiv.org/abs/2601.14264
Abstract:

Large language models (LLMs) are used as “digital twins” to replace human respondents, yet their psychometric comparability to humans is uncertain. We propose a construct-validity framework spanning construct representation and the nomological net, benchmarking digital twins against human gold standards across models, tasks and testing how person-specific inputs shape performance. Across studies, digital twins achieved high population-level accuracy and strong within-participant profile correlations, alongside attenuated item-level correlations. In word association tests, LLM-based networks show small-world structure and theory-consistent communities similar to humans, yet diverge lexically and in local structure. In decision-making and contextualized tasks, digital twins under-reproduce heuristic biases, showing normative rationality, compressed variance and limited sensitivity to temporal information. Feature-rich digital twins improve Big Five Personality prediction, but their personality networks show only configural invariance and do not achieve metric invariance. In more applied free-text tasks, feature-rich digital twins better match human narratives, but linguistic differences persist. Together, these results indicate that feature-rich conditioning enhances validity but does not resolve systematic divergences in psychometric comparability. Future work should therefore prioritize delineating the effective boundaries of digital twins, establishing the precise contexts in which they function as reliable proxies for human cognition and behavior.

LLM 관련 주요 논문 - 2026-01-23

1. Structured Hints for Sample-Efficient Lean Theorem Proving

2. LLM Prompt Evaluation for Educational Applications

3. Multimodal Climate Disinformation Detection: Integrating Vision-Language Models with External Knowledge Sources

4. Controlling Long-Horizon Behavior in Language Model Agents with Explicit State Dynamics

5. Grounding Large Language Models in Reaction Knowledge Graphs for Synthesis Retrieval

6. Deja Vu in Plots: Leveraging Cross-Session Evidence with Retrieval-Augmented LLMs for Live Streaming Risk Assessment

7. ErrorMap and ErrorAtlas: Charting the Failure Landscape of Large Language Models

8. Inference-Time Scaling of Verification: Self-Evolving Deep Research Agents via Test-Time Rubric-Guided Verification

9. VitalDiagnosis: AI-Driven Ecosystem for 24/7 Vital Monitoring and Chronic Disease Management

10. Agentic Confidence Calibration

11. Tabular Incremental Inference

12. AgentSM: Semantic Memory for Agentic Text-to-SQL

13. Improving Methodologies for LLM Evaluations Across Global Languages

14. From Passive Metric to Active Signal: The Evolving Role of Uncertainty Quantification in Large Language Models

15. Predictive Coding and Information Bottleneck for Hallucination Detection in Large Language Models

16. CogToM: A Comprehensive Theory of Mind Benchmark inspired by Human Cognition for Large Language Models

17. Autonomous Business System via Neuro-symbolic AI

18. TransportAgents: a multi-agents LLM framework for traffic accident severity prediction

19. Tracking the Limits of Knowledge Propagation: How LLMs Fail at Multi-Step Reasoning with Conflicting Knowledge

20. Reliability by design: quantifying and eliminating fabrication risk in LLMs. From generative to consultative AI: a comparative analysis in the legal domain and lessons for high-stakes knowledge bases

21. Not Your Typical Sycophant: The Elusive Nature of Sycophancy in Large Language Models

22. Beyond Prompting: Efficient and Robust Contextual Biasing for Speech LLMs via Logit-Space Integration (LOGIC)

23. Prometheus Mind: Retrofitting Memory to Frozen Language Models

24. Replayable Financial Agents: A Determinism-Faithfulness Assurance Harness for Tool-Using LLM Agents

25. The Paradigm Shift: A Comprehensive Survey on Large Vision Language Models for Multimodal Fake News Detection

26. Aeon: High-Performance Neuro-Symbolic Memory Management for Long-Horizon LLM Agents

27. Uncovering Latent Bias in LLM-Based Emergency Department Triage Through Proxy Variables

28. Gated Sparse Attention: Combining Computational Efficiency with Training Stability for Long-Context Language Models

29. LLM-in-Sandbox Elicits General Agentic Intelligence

30. Learning to Discover at Test Time

31. Replicating Human Motivated Reasoning Studies with LLMs

32. Improving Training Efficiency and Reducing Maintenance Costs via Language Specific Model Merging

33. Sawtooth Wavefront Reordering: Enhanced CuTile FlashAttention on NVIDIA GB10

34. PhysicsMind: Sim and Real Mechanics Benchmarking for Physical Reasoning and Prediction in Foundational VLMs and World Models

35. MMGRid: Navigating Temporal-aware and Cross-domain Generative Recommendation via Model Merging

36. TeNet: Text-to-Network for Compact Policy Synthesis

37. Introducing the Generative Application Firewall (GAF)

38. Virtual Traffic Police: Large Language Model-Augmented Traffic Signal Control for Unforeseen Incidents

39. VideoThinker: Building Agentic VideoLLMs with LLM-Guided Tool Reasoning

40. CoNRec: Context-Discerning Negative Recommendation with LLMs

41. FlexLLM: Composable HLS Library for Flexible Hybrid LLM Accelerator Design

42. Beyond Visual Safety: Jailbreaking Multimodal Large Language Models for Harmful Image Generation via Semantic-Agnostic Inputs

43. FARM: Field-Aware Resolution Model for Intelligent Trigger-Action Automation

44. Connect the Dots: Knowledge Graph-Guided Crawler Attack on Retrieval-Augmented Generation Systems

45. TempoNet: Learning Realistic Communication and Timing Patterns for Network Traffic Simulation

46. Event-VStream: Event-Driven Real-Time Understanding for Long Video Streams

47. Robust Tool Use via Fission-GRPO: Learning to Recover from Execution Errors

48. DeepASMR: LLM-Based Zero-Shot ASMR Speech Generation for Anyone of Any Voice

49. Data-Free Privacy-Preserving for LLMs via Model Inversion and Selective Unlearning

50. Parallelism and Generation Order in Masked Diffusion Language Models: Limits Today, Potential Tomorrow

51. MapViT: A Two-Stage ViT-Based Framework for Real-Time Radio Quality Map Prediction in Dynamic Environments

52. VIOLA: Towards Video In-Context Learning with Minimal Annotations

53. Multi-Persona Thinking for Bias Mitigation in Large Language Models

54. The Rise of Large Language Models and the Direction and Impact of US Federal Research Funding

55. Martingale Foresight Sampling: A Principled Approach to Inference-Time LLM Decoding

56. Benchmarking LLMs for Pairwise Causal Discovery in Biomedical and Multi-Domain Contexts

57. Chunking, Retrieval, and Re-ranking: An Empirical Evaluation of RAG Architectures for Policy Document Question Answering

58. CURE: Curriculum-guided Multi-task Training for Reliable Anatomy Grounded Report Generation

59. Beyond Fixed Psychological Personas: State Beats Trait, but Language Models are State-Blind

60. Improving MoE Compute Efficiency by Composing Weight and Data Sparsity

61. Q-Probe: Scaling Image Quality Assessment to High Resolution via Context-Aware Agentic Probing

62. Abusive music and song transformation using GenAI and LLMs

63. Lost in Transcription: How Speech-to-Text Errors Derail Code Understanding

64. ToolCaching: Towards Efficient Caching for LLM Tool-calling

65. No Reliable Evidence of Self-Reported Sentience in Small Large Language Models

66. Empowering LLMs for Structure-Based Drug Design via Exploration-Augmented Latent Inference

67. RECAP: A Resource-Efficient Method for Adversarial Prompting in Large Language Models

68. ICPO: Illocution-Calibrated Policy Optimization for Multi-Turn Conversation

69. Large Language Models as Simulative Agents for Neurodivergent Adult Psychometric Profiles

70. Do people expect different behavior from large language models acting on their behalf? Evidence from norm elicitations in two canonical economic games

71. When Generative AI Meets Extended Reality: Enabling Scalable and Natural Interactions

72. Can We Trust LLM Detectors?

73. Entropy-Tree: Tree-Based Decoding with Entropy-Guided Exploration

74. Agentic Persona Control and Task State Tracking for Realistic User Simulation in Interactive Scenarios

75. LLM-based Multimodal Feedback Produces Equivalent Learning and Better Student Perceptions than Educator Feedback

76. Psychometric Comparability of LLM-Based Digital Twins