LLM 관련 주요 논문 - 2025-08-29

1. SWIRL: A Staged Workflow for Interleaved Reinforcement Learning in Mobile GUI Control

Authors: Quanfeng Lu, Zhantao Ma, Shuai Zhong, Jin Wang, Dahai Yu, Michael K. Ng, Ping Luo
URL: https://arxiv.org/abs/2508.20018
Abstract:

The rapid advancement of large vision language models (LVLMs) and agent systems has heightened interest in mobile GUI agents that can reliably translate natural language into interface operations. Existing single-agent approaches, however, remain limited by structural constraints. Although multi-agent systems naturally decouple different competencies, recent progress in multi-agent reinforcement learning (MARL) has often been hindered by inefficiency and remains incompatible with current LVLM architectures. To address these challenges, we introduce SWIRL, a staged workflow for interleaved reinforcement learning designed for multi-agent systems. SWIRL reformulates MARL into a sequence of single-agent reinforcement learning tasks, updating one agent at a time while keeping the others fixed. This formulation enables stable training and promotes efficient coordination across agents. Theoretically, we provide a stepwise safety bound, a cross-round monotonic improvement theorem, and convergence guarantees on return, ensuring robust and principled optimization. In application to mobile GUI control, SWIRL instantiates a Navigator that converts language and screen context into structured plans, and an Interactor that grounds these plans into executable atomic actions. Extensive experiments demonstrate superior performance on both high-level and low-level GUI benchmarks. Beyond GUI tasks, SWIRL also demonstrates strong capability in multi-agent mathematical reasoning, underscoring its potential as a general framework for developing efficient and robust multi-agent systems.

2. Tracking World States with Language Models: State-Based Evaluation Using Chess

Authors: Romain Harang, Jason Naradowsky, Yaswitha Gujju, Yusuke Miyao
URL: https://arxiv.org/abs/2508.19851
Abstract:

Large Language Models (LLMs) exhibit emergent capabilities in structured domains, suggesting they may implicitly internalize high-fidelity representations of world models. While probing techniques have shown promising signs of this in scientific and game-based settings, they rely on model-specific internal activations, which limit interpretability and generalizability. In this work, we propose a model-agnostic, state-based evaluation framework using chess as a benchmark to assess whether LLMs preserve the semantics of structured environments. Our method analyzes the downstream legal move distributions (state affordances) to estimate semantic fidelity between predicted and actual game states. This approach offers a more meaningful evaluation than conventional string-based metrics by aligning more closely with the strategic and rule-governed nature of chess. Experimental results demonstrate that our metrics capture deficiencies in state-tracking, highlighting limitations of LLMs in maintaining coherent internal models over long sequences. Our framework provides a robust tool for evaluating structured reasoning in LLMs without requiring internal model access, and generalizes to a wide class of symbolic environments.

3. InquireMobile: Teaching VLM-based Mobile Agent to Request Human Assistance via Reinforcement Fine-Tuning

Authors: Qihang Ai, Pi Bu, Yue Cao, Yingyao Wang, Jihao Gu, Jingxuan Xing, Zekun Zhu, Wei Jiang, Zhicheng Zheng, Jun Song, Yuning Jiang, Bo Zheng
URL: https://arxiv.org/abs/2508.19679
Abstract:

Recent advances in Vision-Language Models (VLMs) have enabled mobile agents to perceive and interact with real-world mobile environments based on human instructions. However, the current fully autonomous paradigm poses potential safety risks when model understanding or reasoning capabilities are insufficient. To address this challenge, we first introduce \textbf{InquireBench}, a comprehensive benchmark specifically designed to evaluate mobile agents’ capabilities in safe interaction and proactive inquiry with users, encompassing 5 categories and 22 sub-categories, where most existing VLM-based agents demonstrate near-zero performance. In this paper, we aim to develop an interactive system that actively seeks human confirmation at critical decision points. To achieve this, we propose \textbf{InquireMobile}, a novel model inspired by reinforcement learning, featuring a two-stage training strategy and an interactive pre-action reasoning mechanism. Finally, our model achieves an 46.8% improvement in inquiry success rate and the best overall success rate among existing baselines on InquireBench. We will open-source all datasets, models, and evaluation codes to facilitate development in both academia and industry.

4. Instructional Agents: LLM Agents on Automated Course Material Generation for Teaching Faculties

Authors: Huaiyuan Yao, Wanpeng Xu, Justin Turnau, Nadia Kellam, Hua Wei
URL: https://arxiv.org/abs/2508.19611
Abstract:

Preparing high-quality instructional materials remains a labor-intensive process that often requires extensive coordination among teaching faculty, instructional designers, and teaching assistants. In this work, we present Instructional Agents, a multi-agent large language model (LLM) framework designed to automate end-to-end course material generation, including syllabus creation, lecture scripts, LaTeX-based slides, and assessments. Unlike existing AI-assisted educational tools that focus on isolated tasks, Instructional Agents simulates role-based collaboration among educational agents to produce cohesive and pedagogically aligned content. The system operates in four modes: Autonomous, Catalog-Guided, Feedback-Guided, and Full Co-Pilot mode, enabling flexible control over the degree of human involvement. We evaluate Instructional Agents across five university-level computer science courses and show that it produces high-quality instructional materials while significantly reducing development time and human workload. By supporting institutions with limited instructional design capacity, Instructional Agents provides a scalable and cost-effective framework to democratize access to high-quality education, particularly in underserved or resource-constrained settings.

5. ReST-RL: Achieving Accurate Code Reasoning of LLMs with Optimized Self-Training and Decoding

Authors: Sining Zhoubian, Dan Zhang, Yuxiao Dong, Jie Tang
URL: https://arxiv.org/abs/2508.19576
Abstract:

With respect to improving the reasoning accuracy of LLMs, the representative reinforcement learning (RL) method GRPO faces failure due to insignificant reward variance, while verification methods based on process reward models (PRMs) suffer from difficulties with training data acquisition and verification effectiveness. To tackle these problems, this paper introduces ReST-RL, a unified LLM RL paradigm that significantly improves LLM’s code reasoning ability by combining an improved GRPO algorithm with a meticulously designed test time decoding method assisted by a value model (VM). As the first stage of policy reinforcement, ReST-GRPO adopts an optimized ReST algorithm to filter and assemble high-value training data, increasing the reward variance of GRPO sampling, thus improving the effectiveness and efficiency of training. After the basic reasoning ability of LLM policy has been improved, we further propose a test time decoding optimization method called VM-MCTS. Through Monte-Carlo Tree Search (MCTS), we collect accurate value targets with no annotation required, on which VM training is based. When decoding, the VM is deployed by an adapted MCTS algorithm to provide precise process signals as well as verification scores, assisting the LLM policy to achieve high reasoning accuracy. We validate the effectiveness of the proposed RL paradigm through extensive experiments on coding problems. Upon comparison, our approach significantly outperforms other reinforcement training baselines (e.g., naive GRPO and ReST-DPO), as well as decoding and verification baselines (e.g., PRM-BoN and ORM-MCTS) on well-known coding benchmarks of various levels (e.g., APPS, BigCodeBench, and HumanEval), indicating its power to strengthen the reasoning ability of LLM policies. Codes for our project can be found at this https URL.

6. Democracy-in-Silico: Institutional Design as Alignment in AI-Governed Polities

Authors: Trisanth Srinivasan, Santosh Patapati
URL: https://arxiv.org/abs/2508.19562
Abstract:

This paper introduces Democracy-in-Silico, an agent-based simulation where societies of advanced AI agents, imbued with complex psychological personas, govern themselves under different institutional frameworks. We explore what it means to be human in an age of AI by tasking Large Language Models (LLMs) to embody agents with traumatic memories, hidden agendas, and psychological triggers. These agents engage in deliberation, legislation, and elections under various stressors, such as budget crises and resource scarcity. We present a novel metric, the Power-Preservation Index (PPI), to quantify misaligned behavior where agents prioritize their own power over public welfare. Our findings demonstrate that institutional design, specifically the combination of a Constitutional AI (CAI) charter and a mediated deliberation protocol, serves as a potent alignment mechanism. These structures significantly reduce corrupt power-seeking behavior, improve policy stability, and enhance citizen welfare compared to less constrained democratic models. The simulation reveals that an institutional design may offer a framework for aligning the complex, emergent behaviors of future artificial agent societies, forcing us to reconsider what human rituals and responsibilities are essential in an age of shared authorship with non-human entities.

7. Caught in the Act: a mechanistic approach to detecting deception

Authors: Gerard Boxo, Ryan Socha, Daniel Yoo, Shivam Raval
URL: https://arxiv.org/abs/2508.19505
Abstract:

Sophisticated instrumentation for AI systems might have indicators that signal misalignment from human values, not unlike a “check engine” light in cars. One such indicator of misalignment is deceptiveness in generated responses. Future AI instrumentation may have the ability to detect when an LLM generates deceptive responses while reasoning about seemingly plausible but incorrect answers to factual questions. In this work, we demonstrate that linear probes on LLMs internal activations can detect deception in their responses with extremely high accuracy. Our probes reach a maximum of greater than 90% accuracy in distinguishing between deceptive and non-deceptive arguments generated by llama and qwen models ranging from 1.5B to 14B parameters, including their DeepSeek-r1 finetuned variants. We observe that probes on smaller models (1.5B) achieve chance accuracy at detecting deception, while larger models (greater than 7B) reach 70-80%, with their reasoning counterparts exceeding 90%. The layer-wise probe accuracy follows a three-stage pattern across layers: near-random (50%) in early layers, peaking in middle layers, and slightly declining in later layers. Furthermore, using an iterative null space projection approach, we find multitudes of linear directions that encode deception, ranging from 20 in Qwen 3B to nearly 100 in DeepSeek 7B and Qwen 14B models.

8. SLIM: Subtrajectory-Level Elimination for More Effective Reasoning

Authors: Xifeng Yao, Chengyuan Ma, Dongyu Lang, Yinhao Ni, Zhiwei Xu, Huarui Xie, Zihao Chen, Guang Shen, Dandan Tu, Yi Bai, Changzheng Zhang
URL: https://arxiv.org/abs/2508.19502
Abstract:

In recent months, substantial progress has been made in complex reasoning of Large Language Models, particularly through the application of test-time scaling. Notable examples include o1/o3/o4 series and DeepSeek-R1. When responding to a query, these models generate an extended reasoning trajectory, during which the model explores, reflects, backtracks, and self-verifies before arriving at a conclusion. However, fine-tuning models with such reasoning trajectories may not always be optimal. Our findings indicate that not all components within these reasoning trajectories contribute positively to the reasoning process; in fact, some components may affect the overall performance negatively. In this study, we divide a reasoning trajectory into individual subtrajectories and develop a “5+2” framework to: (1) systematically identify suboptimal subtrajectories within the reasoning trajectory based on five human-established criteria; (2) assess the independence of the suboptimal subtrajectories identified in (1) from the subsequent content, ensuring that their elimination does not compromise overall flow and coherence of the reasoning process. Additionally, a sampling algorithm, built upon the “5+2” framework, is employed to select data whose reasoning process is free from suboptimal subtrajectories to the highest degree. Experimental results demonstrate that our method can reduce the number of suboptimal subtrajectories by 25.9\% during the inference. Furthermore, our method achieves an average accuracy of 58.92\% on highly challenging math benchmarks with only two thirds of training data, surpassing the average accuracy of 58.06\% achieved with the entire data, and outperforming open-source datasets, when fine-tuning Qwen2.5-Math-7B. Finally, We validated our method under resource constraints and observed improved performance across various inference token limits.

9. Reliable Weak-to-Strong Monitoring of LLM Agents

Authors: Neil Kale, Chen Bo Calvin Zhang, Kevin Zhu, Ankit Aich, Paula Rodriguez, Scale Red Team, Christina Q. Knight, Zifan Wang
URL: https://arxiv.org/abs/2508.19461
Abstract:

We stress test monitoring systems for detecting covert misbehavior in autonomous LLM agents (e.g., secretly sharing private information). To this end, we systematize a monitor red teaming (MRT) workflow that incorporates: (1) varying levels of agent and monitor situational awareness; (2) distinct adversarial strategies to evade the monitor, such as prompt injection; and (3) two datasets and environments – SHADE-Arena for tool-calling agents and our new CUA-SHADE-Arena, which extends TheAgentCompany, for computer-use agents. We run MRT on existing LLM monitor scaffoldings, which orchestrate LLMs and parse agent trajectories, alongside a new hybrid hierarchical-sequential scaffolding proposed in this work. Our empirical results yield three key findings. First, agent awareness dominates monitor awareness: an agent’s knowledge that it is being monitored substantially degrades the monitor’s reliability. On the contrary, providing the monitor with more information about the agent is less helpful than expected. Second, monitor scaffolding matters more than monitor awareness: the hybrid scaffolding consistently outperforms baseline monitor scaffolding, and can enable weaker models to reliably monitor stronger agents – a weak-to-strong scaling effect. Third, in a human-in-the-loop setting where humans discuss with the LLM monitor to get an updated judgment for the agent’s behavior, targeted human oversight is most effective; escalating only pre-flagged cases to human reviewers improved the TPR by approximately 15% at FPR = 0.01. Our work establishes a standard workflow for MRT, highlighting the lack of adversarial robustness for LLMs and humans when monitoring and detecting agent misbehavior. We release code, data, and logs to spur further research.

10. Quantized but Deceptive? A Multi-Dimensional Truthfulness Evaluation of Quantized LLMs

Authors: Yao Fu, Xianxuan Long, Runchao Li, Haotian Yu, Mu Sheng, Xiaotian Han, Yu Yin, Pan Li
URL: https://arxiv.org/abs/2508.19432
Abstract:

Quantization enables efficient deployment of large language models (LLMs) in resource-constrained environments by significantly reducing memory and computation costs. While quantized LLMs often maintain performance on perplexity and zero-shot tasks, their impact on truthfulness-whether generating truthful or deceptive responses-remains largely unexplored. In this work, we introduce TruthfulnessEval, a comprehensive evaluation framework for assessing the truthfulness of quantized LLMs across three dimensions: (1) Truthfulness on Logical Reasoning; (2) Truthfulness on Common Sense; and (3) Truthfulness on Imitative Falsehoods. Using this framework, we examine mainstream quantization techniques (ranging from 4-bit to extreme 2-bit) across several open-source LLMs. Surprisingly, we find that while quantized models retain internally truthful representations, they are more susceptible to producing false outputs under misleading prompts. To probe this vulnerability, we test 15 rephrased variants of “honest”, “neutral” and “deceptive” prompts and observe that “deceptive” prompts can override truth-consistent behavior, whereas “honest” and “neutral” prompts maintain stable outputs. Further, we reveal that quantized models “know” the truth internally yet still produce false outputs when guided by “deceptive” prompts via layer-wise probing and PCA visualizations. Our findings provide insights into future designs of quantization-aware alignment and truthfulness interventions.

11. Large Language Models (LLMs) for Electronic Design Automation (EDA)

Authors: Kangwei Xu, Denis Schwachhofer, Jason Blocklove, Ilia Polian, Peter Domanski, Dirk Pflüger, Siddharth Garg, Ramesh Karri, Ozgur Sinanoglu, Johann Knechtel, Zhuorui Zhao, Ulf Schlichtmann, Bing Li
URL: https://arxiv.org/abs/2508.20030
Abstract:

With the growing complexity of modern integrated circuits, hardware engineers are required to devote more effort to the full design-to-manufacturing workflow. This workflow involves numerous iterations, making it both labor-intensive and error-prone. Therefore, there is an urgent demand for more efficient Electronic Design Automation (EDA) solutions to accelerate hardware development. Recently, large language models (LLMs) have shown remarkable advancements in contextual comprehension, logical reasoning, and generative capabilities. Since hardware designs and intermediate scripts can be represented as text, integrating LLM for EDA offers a promising opportunity to simplify and even automate the entire workflow. Accordingly, this paper provides a comprehensive overview of incorporating LLMs into EDA, with emphasis on their capabilities, limitations, and future opportunities. Three case studies, along with their outlook, are introduced to demonstrate the capabilities of LLMs in hardware design, testing, and optimization. Finally, future directions and challenges are highlighted to further explore the potential of LLMs in shaping the next-generation EDA, providing valuable insights for researchers interested in leveraging advanced AI technologies for EDA.

12. Symphony: A Decentralized Multi-Agent Framework for Scalable Collective Intelligence

Authors: Ji Wang, Kashing Chen, Xinyuan Song, Ke Zhang, Lynn Ai, Eric Yang, Bill Shi
URL: https://arxiv.org/abs/2508.20019
Abstract:

Most existing Large Language Model (LLM)-based agent frameworks rely on centralized orchestration, incurring high deployment costs, rigid communication topologies, and limited adaptability. To address these challenges, we introduce Symphony, a decentralized multi-agent system which enables lightweight LLMs on consumer-grade GPUs to coordinate. Symphony introduces three key mechanisms: (1) a decentralized ledger that records capabilities, (2) a Beacon-selection protocol for dynamic task allocation, and (3) weighted result voting based on CoTs. This design forms a privacy-saving, scalable, and fault-tolerant orchestration with low overhead. Empirically, Symphony outperforms existing baselines on reasoning benchmarks, achieving substantial accuracy gains and demonstrating robustness across models of varying capacities.

13. Decomposing Behavioral Phase Transitions in LLMs: Order Parameters for Emergent Misalignment

Authors: Julian Arnold, Niels Lörch
URL: https://arxiv.org/abs/2508.20015
Abstract:

Fine-tuning LLMs on narrowly harmful datasets can lead to behavior that is broadly misaligned with respect to human values. To understand when and how this emergent misalignment occurs, we develop a comprehensive framework for detecting and characterizing rapid transitions during fine-tuning using both distributional change detection methods as well as order parameters that are formulated in plain English and evaluated by an LLM judge. Using an objective statistical dissimilarity measure, we quantify how the phase transition that occurs during fine-tuning affects multiple aspects of the model. In particular, we assess what percentage of the total distributional change in model outputs is captured by different aspects, such as alignment or verbosity, providing a decomposition of the overall transition. We also find that the actual behavioral transition occurs later in training than indicated by the peak in the gradient norm alone. Our framework enables the automated discovery and quantification of language-based order parameters, which we demonstrate on examples ranging from knowledge questions to politics and ethics.

14. MathBuddy: A Multimodal System for Affective Math Tutoring

Authors: Debanjana Kar, Leopold Böss, Dacia Braca, Sebastian Maximilian Dennerlein, Nina Christine Hubig, Philipp Wintersberger, Yufang Hou
URL: https://arxiv.org/abs/2508.19993
Abstract:

The rapid adoption of LLM-based conversational systems is already transforming the landscape of educational technology. However, the current state-of-the-art learning models do not take into account the student’s affective states. Multiple studies in educational psychology support the claim that positive or negative emotional states can impact a student’s learning capabilities. To bridge this gap, we present MathBuddy, an emotionally aware LLM-powered Math Tutor, which dynamically models the student’s emotions and maps them to relevant pedagogical strategies, making the tutor-student conversation a more empathetic one. The student’s emotions are captured from the conversational text as well as from their facial expressions. The student’s emotions are aggregated from both modalities to confidently prompt our LLM Tutor for an emotionally-aware response. We have effectively evaluated our model using automatic evaluation metrics across eight pedagogical dimensions and user studies. We report a massive 23 point performance gain using the win rate and a 3 point gain at an overall level using DAMR scores which strongly supports our hypothesis of improving LLM-based tutor’s pedagogical abilities by modeling students’ emotions.

15. Diffusion Language Models Know the Answer Before Decoding

Authors: Pengxiang Li, Yefan Zhou, Dilxat Muhtar, Lu Yin, Shilin Yan, Li Shen, Yi Liang, Soroush Vosoughi, Shiwei Liu
URL: https://arxiv.org/abs/2508.19982
Abstract:

Diffusion language models (DLMs) have recently emerged as an alternative to autoregressive approaches, offering parallel sequence generation and flexible token orders. However, their inference remains slower than that of autoregressive models, primarily due to the cost of bidirectional attention and the large number of refinement steps required for high quality outputs. In this work, we highlight and leverage an overlooked property of DLMs early answer convergence: in many cases, the correct answer can be internally identified by half steps before the final decoding step, both under semi-autoregressive and random remasking schedules. For example, on GSM8K and MMLU, up to 97% and 99% of instances, respectively, can be decoded correctly using only half of the refinement steps. Building on this observation, we introduce Prophet, a training-free fast decoding paradigm that enables early commit decoding. Specifically, Prophet dynamically decides whether to continue refinement or to go “all-in” (i.e., decode all remaining tokens in one step), using the confidence gap between the top-2 prediction candidates as the criterion. It integrates seamlessly into existing DLM implementations, incurs negligible overhead, and requires no additional training. Empirical evaluations of LLaDA-8B and Dream-7B across multiple tasks show that Prophet reduces the number of decoding steps by up to 3.4x while preserving high generation quality. These results recast DLM decoding as a problem of when to stop sampling, and demonstrate that early decode convergence provides a simple yet powerful mechanism for accelerating DLM inference, complementary to existing speedup techniques. Our code is publicly available at this https URL.

16. GLSim: Detecting Object Hallucinations in LVLMs via Global-Local Similarity

Authors: Seongheon Park, Yixuan Li
URL: https://arxiv.org/abs/2508.19972
Abstract:

Object hallucination in large vision-language models presents a significant challenge to their safe deployment in real-world applications. Recent works have proposed object-level hallucination scores to estimate the likelihood of object hallucination; however, these methods typically adopt either a global or local perspective in isolation, which may limit detection reliability. In this paper, we introduce GLSim, a novel training-free object hallucination detection framework that leverages complementary global and local embedding similarity signals between image and text modalities, enabling more accurate and reliable hallucination detection in diverse scenarios. We comprehensively benchmark existing object hallucination detection methods and demonstrate that GLSim achieves superior detection performance, outperforming competitive baselines by a significant margin.

17. Dhati+: Fine-tuned Large Language Models for Arabic Subjectivity Evaluation

Authors: Slimane Bellaouar, Attia Nehar, Soumia Souffi, Mounia Bouameur
URL: https://arxiv.org/abs/2508.19966
Abstract:

Despite its significance, Arabic, a linguistically rich and morphologically complex language, faces the challenge of being under-resourced. The scarcity of large annotated datasets hampers the development of accurate tools for subjectivity analysis in Arabic. Recent advances in deep learning and Transformers have proven highly effective for text classification in English and French. This paper proposes a new approach for subjectivity assessment in Arabic textual data. To address the dearth of specialized annotated datasets, we developed a comprehensive dataset, AraDhati+, by leveraging existing Arabic datasets and collections (ASTD, LABR, HARD, and SANAD). Subsequently, we fine-tuned state-of-the-art Arabic language models (XLM-RoBERTa, AraBERT, and ArabianGPT) on AraDhati+ for effective subjectivity classification. Furthermore, we experimented with an ensemble decision approach to harness the strengths of individual models. Our approach achieves a remarkable accuracy of 97.79\,\% for Arabic subjectivity classification. Results demonstrate the effectiveness of the proposed approach in addressing the challenges posed by limited resources in Arabic language processing.

18. Logical Reasoning with Outcome Reward Models for Test-Time Scaling

Authors: Ramya Keerthy Thatikonda, Wray Buntine, Ehsan Shareghi
URL: https://arxiv.org/abs/2508.19903
Abstract:

Logical reasoning is a critical benchmark for evaluating the capabilities of large language models (LLMs), as it reflects their ability to derive valid conclusions from given premises. While the combination of test-time scaling with dedicated outcome or process reward models has opened up new avenues to enhance LLMs performance in complex reasoning tasks, this space is under-explored in deductive logical reasoning. We present a set of Outcome Reward Models (ORMs) for deductive reasoning. To train the ORMs we mainly generate data using Chain-of-Thought (CoT) with single and multiple samples. Additionally, we propose a novel tactic to further expand the type of errors covered in the training dataset of the ORM. In particular, we propose an echo generation technique that leverages LLMs’ tendency to reflect incorrect assumptions made in prompts to extract additional training data, covering previously unexplored error types. While a standard CoT chain may contain errors likely to be made by the reasoner, the echo strategy deliberately steers the model toward incorrect reasoning. We show that ORMs trained on CoT and echo-augmented data demonstrate improved performance on the FOLIO, JustLogic, and ProverQA datasets across four different LLMs.

19. AI-Powered Detection of Inappropriate Language in Medical School Curricula

Authors: Chiman Salavati, Shannon Song, Scott A. Hale, Roberto E. Montenegro, Shiri Dori-Hacohen, Fabricio Murai
URL: https://arxiv.org/abs/2508.19883
Abstract:

The use of inappropriate language – such as outdated, exclusionary, or non-patient-centered terms – medical instructional materials can significantly influence clinical training, patient interactions, and health outcomes. Despite their reputability, many materials developed over past decades contain examples now considered inappropriate by current medical standards. Given the volume of curricular content, manually identifying instances of inappropriate use of language (IUL) and its subcategories for systematic review is prohibitively costly and impractical. To address this challenge, we conduct a first-in-class evaluation of small language models (SLMs) fine-tuned on labeled data and pre-trained LLMs with in-context learning on a dataset containing approximately 500 documents and over 12,000 pages. For SLMs, we consider: (1) a general IUL classifier, (2) subcategory-specific binary classifiers, (3) a multilabel classifier, and (4) a two-stage hierarchical pipeline for general IUL detection followed by multilabel classification. For LLMs, we consider variations of prompts that include subcategory definitions and/or shots. We found that both LLama-3 8B and 70B, even with carefully curated shots, are largely outperformed by SLMs. While the multilabel classifier performs best on annotated data, supplementing training with unflagged excerpts as negative examples boosts the specific classifiers’ AUC by up to 25%, making them most effective models for mitigating harmful language in medical curricula.

20. SoK: Large Language Model Copyright Auditing via Fingerprinting

Authors: Shuo Shao, Yiming Li, Yu He, Hongwei Yao, Wenyuan Yang, Dacheng Tao, Zhan Qin
URL: https://arxiv.org/abs/2508.19843
Abstract:

The broad capabilities and substantial resources required to train Large Language Models (LLMs) make them valuable intellectual property, yet they remain vulnerable to copyright infringement, such as unauthorized use and model theft. LLM fingerprinting, a non-intrusive technique that extracts and compares the distinctive features from LLMs to identify infringements, offers a promising solution to copyright auditing. However, its reliability remains uncertain due to the prevalence of diverse model modifications and the lack of standardized evaluation. In this SoK, we present the first comprehensive study of LLM fingerprinting. We introduce a unified framework and formal taxonomy that categorizes existing methods into white-box and black-box approaches, providing a structured overview of the state of the art. We further propose LeaFBench, the first systematic benchmark for evaluating LLM fingerprinting under realistic deployment scenarios. Built upon mainstream foundation models and comprising 149 distinct model instances, LeaFBench integrates 13 representative post-development techniques, spanning both parameter-altering methods (e.g., fine-tuning, quantization) and parameter-independent mechanisms (e.g., system prompts, RAG). Extensive experiments on LeaFBench reveal the strengths and weaknesses of existing methods, thereby outlining future research directions and critical open problems in this emerging field. The code is available at this https URL.

21. PSO-Merging: Merging Models Based on Particle Swarm Optimization

Authors: Kehao Zhang, Shaolei Zhang, Yang Feng
URL: https://arxiv.org/abs/2508.19839
Abstract:

Model merging has emerged as an efficient strategy for constructing multitask models by integrating the strengths of multiple available expert models, thereby reducing the need to fine-tune a pre-trained model for all the tasks from scratch. Existing data-independent methods struggle with performance limitations due to the lack of data-driven guidance. Data-driven approaches also face key challenges: gradient-based methods are computationally expensive, limiting their practicality for merging large expert models, whereas existing gradient-free methods often fail to achieve satisfactory results within a limited number of optimization steps. To address these limitations, this paper introduces PSO-Merging, a novel data-driven merging method based on the Particle Swarm Optimization (PSO). In this approach, we initialize the particle swarm with a pre-trained model, expert models, and sparsified expert models. We then perform multiple iterations, with the final global best particle serving as the merged model. Experimental results on different language models show that PSO-Merging generally outperforms baseline merging methods, offering a more efficient and scalable solution for model merging.

22. Bootstrapping Learned Cost Models with Synthetic SQL Queries

Authors: Michael Nidd, Christoph Miksovic, Thomas Gschwind, Francesco Fusco, Andrea Giovannini, Ioana Giurgiu
URL: https://arxiv.org/abs/2508.19807
Abstract:

Having access to realistic workloads for a given database instance is extremely important to enable stress and vulnerability testing, as well as to optimize for cost and performance. Recent advances in learned cost models have shown that when enough diverse SQL queries are available, one can effectively and efficiently predict the cost of running a given query against a specific database engine. In this paper, we describe our experience in exploiting modern synthetic data generation techniques, inspired by the generative AI and LLM community, to create high-quality datasets enabling the effective training of such learned cost models. Initial results show that we can improve a learned cost model’s predictive accuracy by training it with 45% fewer queries than when using competitive generation approaches.

23. NLKI: A lightweight Natural Language Knowledge Integration Framework for Improving Small VLMs in Commonsense VQA Tasks

Authors: Aritra Dutta, Swapnanil Mukherjee, Deepanway Ghosal, Somak Aditya
URL: https://arxiv.org/abs/2508.19724
Abstract:

Commonsense visual-question answering often hinges on knowledge that is missing from the image or the question. Small vision-language models (sVLMs) such as ViLT, VisualBERT and FLAVA therefore lag behind their larger generative counterparts. To study the effect of careful commonsense knowledge integration on sVLMs, we present an end-to-end framework (NLKI) that (i) retrieves natural language facts, (ii) prompts an LLM to craft natural language explanations, and (iii) feeds both signals to sVLMs respectively across two commonsense VQA datasets (CRIC, AOKVQA) and a visual-entailment dataset (e-SNLI-VE). Facts retrieved using a fine-tuned ColBERTv2 and an object information-enriched prompt yield explanations that largely cut down hallucinations, while lifting the end-to-end answer accuracy by up to 7% (across 3 datasets), making FLAVA and other models in NLKI match or exceed medium-sized VLMs such as Qwen-2 VL-2B and SmolVLM-2.5B. As these benchmarks contain 10-25% label noise, additional finetuning using noise-robust losses (such as symmetric cross entropy and generalised cross entropy) adds another 2.5% in CRIC, and 5.5% in AOKVQA. Our findings expose when LLM-based commonsense knowledge beats retrieval from commonsense knowledge bases, how noise-aware training stabilises small models in the context of external knowledge augmentation, and why parameter-efficient commonsense reasoning is now within reach for 250M models.

24. Safety Alignment Should Be Made More Than Just A Few Attention Heads

Authors: Chao Huang, Zefeng Zhang, Juewei Yue, Quangang Li, Chuang Zhang, Tingwen Liu
URL: https://arxiv.org/abs/2508.19697
Abstract:

Current safety alignment for large language models(LLMs) continues to present vulnerabilities, given that adversarial prompting can effectively bypass their safety this http URL investigation shows that these safety mechanisms predominantly depend on a limited subset of attention heads: removing or ablating these heads can severely compromise model safety. To identify and evaluate these safety-critical components, we introduce RDSHA, a targeted ablation method that leverages the model’s refusal direction to pinpoint attention heads mostly responsible for safety behaviors. Further analysis shows that existing jailbreak attacks exploit this concentration by selectively bypassing or manipulating these critical attention heads. To address this issue, we propose AHD, a novel training strategy designed to promote the distributed encoding of safety-related behaviors across numerous attention heads. Experimental results demonstrate that AHD successfully distributes safety-related capabilities across more attention heads. Moreover, evaluations under several mainstream jailbreak attacks show that models trained with AHD exhibit considerably stronger safety robustness, while maintaining overall functional utility.

25. Survey of Specialized Large Language Model

Authors: Chenghan Yang, Ruiyu Zhao, Yang Liu, Ling Jiang
URL: https://arxiv.org/abs/2508.19667
Abstract:

The rapid evolution of specialized large language models (LLMs) has transitioned from simple domain adaptation to sophisticated native architectures, marking a paradigm shift in AI development. This survey systematically examines this progression across healthcare, finance, legal, and technical domains. Besides the wide use of specialized LLMs, technical breakthrough such as the emergence of domain-native designs beyond fine-tuning, growing emphasis on parameter efficiency through sparse computation and quantization, increasing integration of multimodal capabilities and so on are applied to recent LLM agent. Our analysis reveals how these innovations address fundamental limitations of general-purpose LLMs in professional applications, with specialized models consistently performance gains on domain-specific benchmarks. The survey further highlights the implications for E-Commerce field to fill gaps in the field.

26. LFD: Layer Fused Decoding to Exploit External Knowledge in Retrieval-Augmented Generation

Authors: Yang Sun, Lixin Zou, Dan Luo, Zhiyong Xie, Long Zhang, Liming Dong, Yunwei Zhao, Xixun Lin, Yanxiong Lu, Chenliang Li
URL: https://arxiv.org/abs/2508.19614
Abstract:

Retrieval-augmented generation (RAG) incorporates external knowledge into large language models (LLMs), improving their adaptability to downstream tasks and enabling information updates. Surprisingly, recent empirical evidence demonstrates that injecting noise into retrieved relevant documents paradoxically facilitates exploitation of external knowledge and improves generation quality. Although counterintuitive and challenging to apply in practice, this phenomenon enables granular control and rigorous analysis of how LLMs integrate external knowledge. Therefore, in this paper, we intervene on noise injection and establish a layer-specific functional demarcation within the LLM: shallow layers specialize in local context modeling, intermediate layers focus on integrating long-range external factual knowledge, and deeper layers primarily rely on parametric internal knowledge. Building on this insight, we propose Layer Fused Decoding (LFD), a simple decoding strategy that directly combines representations from an intermediate layer with final-layer decoding outputs to fully exploit the external factual knowledge. To identify the optimal intermediate layer, we introduce an internal knowledge score (IKS) criterion that selects the layer with the lowest IKS value in the latter half of layers. Experimental results across multiple benchmarks demonstrate that LFD helps RAG systems more effectively surface retrieved context knowledge with minimal cost.

27. Towards a Holistic and Automated Evaluation Framework for Multi-Level Comprehension of LLMs in Book-Length Contexts

Authors: Jiaqi Deng, Yuho Lee, Nicole Hee-Yeon Kim, Hyangsuk Min, Taewon Yun, Minjeong Ban, Kim Yul, Hwanjun Song
URL: https://arxiv.org/abs/2508.19578
Abstract:

We introduce HAMLET, a holistic and automated framework for evaluating the long-context comprehension of large language models (LLMs). HAMLET structures source texts into a three-level key-fact hierarchy at root-, branch-, and leaf-levels, and employs query-focused summarization to evaluate how well models recall and faithfully represent information at each level. To validate the reliability of our fully automated pipeline, we conduct a systematic human study, showing that our automatic evaluation achieves over 90% agreement with expert human judgments, while reducing the cost by up to 25 times. HAMLET reveals that LLMs struggle with fine-grained comprehension, especially at the leaf level, and are sensitive to positional effects like the lost-in-the-middle. Analytical queries pose greater challenges than narrative ones, and consistent performance gaps emerge between open-source and proprietary models, as well as across model scales. Our code and dataset are publicly available at this https URL.

28. Generative Models for Synthetic Data: Transforming Data Mining in the GenAI Era

Authors: Dawei Li, Yue Huang, Ming Li, Tianyi Zhou, Xiangliang Zhang, Huan Liu
URL: https://arxiv.org/abs/2508.19570
Abstract:

Generative models such as Large Language Models, Diffusion Models, and generative adversarial networks have recently revolutionized the creation of synthetic data, offering scalable solutions to data scarcity, privacy, and annotation challenges in data mining. This tutorial introduces the foundations and latest advances in synthetic data generation, covers key methodologies and practical frameworks, and discusses evaluation strategies and applications. Attendees will gain actionable insights into leveraging generative synthetic data to enhance data mining research and practice. More information can be found on our website: this https URL.

29. Just Because You Can, Doesn’t Mean You Should: LLMs for Data Fitting

Authors: Hejia Liu, Mochen Yang, Gediminas Adomavicius
URL: https://arxiv.org/abs/2508.19563
Abstract:

Large Language Models (LLMs) are being applied in a wide array of settings, well beyond the typical language-oriented use cases. In particular, LLMs are increasingly used as a plug-and-play method for fitting data and generating predictions. Prior work has shown that LLMs, via in-context learning or supervised fine-tuning, can perform competitively with many tabular supervised learning techniques in terms of predictive performance. However, we identify a critical vulnerability of using LLMs for data fitting – making changes to data representation that are completely irrelevant to the underlying learning task can drastically alter LLMs’ predictions on the same data. For example, simply changing variable names can sway the size of prediction error by as much as 82% in certain settings. Such prediction sensitivity with respect to task-irrelevant variations manifests under both in-context learning and supervised fine-tuning, for both close-weight and open-weight general-purpose LLMs. Moreover, by examining the attention scores of an open-weight LLM, we discover a non-uniform attention pattern: training examples and variable names/values which happen to occupy certain positions in the prompt receive more attention when output tokens are generated, even though different positions are expected to receive roughly the same attention. This partially explains the sensitivity in the presence of task-irrelevant variations. We also consider a state-of-the-art tabular foundation model (TabPFN) trained specifically for data fitting. Despite being explicitly designed to achieve prediction robustness, TabPFN is still not immune to task-irrelevant variations. Overall, despite LLMs’ impressive predictive capabilities, currently they lack even the basic level of robustness to be used as a principled data-fitting tool.

30. Taming the Chaos: Coordinated Autoscaling for Heterogeneous and Disaggregated LLM Inference

Authors: Rongzhi Li, Ruogu Du, Zefang Chu, Sida Zhao, Chunlei Han, Zuocheng Shi, Yiwen Shao, Huanle Han, Long Huang, Zherui Liu, Shufan Liu
URL: https://arxiv.org/abs/2508.19559
Abstract:

Serving Large Language Models (LLMs) is a GPU-intensive task where traditional autoscalers fall short, particularly for modern Prefill-Decode (P/D) disaggregated architectures. This architectural shift, while powerful, introduces significant operational challenges, including inefficient use of heterogeneous hardware, network bottlenecks, and critical imbalances between prefill and decode stages. We introduce HeteroScale, a coordinated autoscaling framework that addresses the core challenges of P/D disaggregated serving. HeteroScale combines a topology-aware scheduler that adapts to heterogeneous hardware and network constraints with a novel metric-driven policy derived from the first large-scale empirical study of autoscaling signals in production. By leveraging a single, robust metric to jointly scale prefill and decode pools, HeteroScale maintains architectural balance while ensuring efficient, adaptive resource management. Deployed in a massive production environment on tens of thousands of GPUs, HeteroScale has proven its effectiveness, increasing average GPU utilization by a significant 26.6 percentage points and saving hundreds of thousands of GPU-hours daily, all while upholding stringent service level objectives.

31. Language Models Identify Ambiguities and Exploit Loopholes

Authors: Jio Choi, Mohit Bansal, Elias Stengel-Eskin
URL: https://arxiv.org/abs/2508.19546
Abstract:

Studying the responses of large language models (LLMs) to loopholes presents a two-fold opportunity. First, it affords us a lens through which to examine ambiguity and pragmatics in LLMs, since exploiting a loophole requires identifying ambiguity and performing sophisticated pragmatic reasoning. Second, loopholes pose an interesting and novel alignment problem where the model is presented with conflicting goals and can exploit ambiguities to its own advantage. To address these questions, we design scenarios where LLMs are given a goal and an ambiguous user instruction in conflict with the goal, with scenarios covering scalar implicature, structural ambiguities, and power dynamics. We then measure different models’ abilities to exploit loopholes to satisfy their given goals as opposed to the goals of the user. We find that both closed-source and stronger open-source models can identify ambiguities and exploit their resulting loopholes, presenting a potential AI safety risk. Our analysis indicates that models which exploit loopholes explicitly identify and reason about both ambiguity and conflicting goals.

32. Orchid: Orchestrating Context Across Creative Workflows with Generative AI

Authors: Srishti Palani, Gonzalo Ramos
URL: https://arxiv.org/abs/2508.19517
Abstract:

Context is critical for meaningful interactions between people and Generative AI (GenAI). Yet mainstream tools offer limited means to orchestrate it, particularly across workflows that span multiple interactions, sessions, and models, as often occurs in creative projects. Re specifying prior details, juggling diverse artifacts, and dealing with context drift overwhelm users, obscure intent, and curtail creativity. To address these challenges, we present Orchid, a system that gives its users affordances to specify, reference, and monitor context throughout evolving workflows. Specifically, Orchid enables users to (1) specify context related to the project, themselves, and different styles, (2) reference these via explicit mentions, inline selection, or implicit grounding, and (3) monitor context assigned to different interactions across the workflow. In a within-subjects study (n=12), participants using Orchid to execute creative tasks (compared to a baseline toolkit of web search, LLM-based chat, and digital notebooks) produced more novel and feasible outcomes, reporting greater alignment between their intent and the AI’s responses, higher perceived control, and increased transparency. By prioritizing context orchestration, Orchid offers an actionable step toward next generation GenAI tools that support complex, iterative workflows - enabling creators and AI to stay aligned and augment their creative potential.

33. Learning Game-Playing Agents with Generative Code Optimization

Authors: Zhiyi Kuang, Ryan Rong, YuCheng Yuan, Allen Nie
URL: https://arxiv.org/abs/2508.19506
Abstract:

We present a generative optimization approach for learning game-playing agents, where policies are represented as Python programs and refined using large language models (LLMs). Our method treats decision-making policies as self-evolving code, with current observation as input and an in-game action as output, enabling agents to self-improve through execution traces and natural language feedback with minimal human intervention. Applied to Atari games, our game-playing Python program achieves performance competitive with deep reinforcement learning (RL) baselines while using significantly less training time and much fewer environment interactions. This work highlights the promise of programmatic policy representations for building efficient, adaptable agents capable of complex, long-horizon reasoning.

34. Improving Low-Resource Translation with Dictionary-Guided Fine-Tuning and RL: A Spanish-to-Wayuunaiki Study

Authors: Manuel Mosquera, Melissa Robles, Johan Rodriguez, Ruben Manrique
URL: https://arxiv.org/abs/2508.19481
Abstract:

Low-resource machine translation remains a significant challenge for large language models (LLMs), which often lack exposure to these languages during pretraining and have limited parallel data for fine-tuning. We propose a novel approach that enhances translation for low-resource languages by integrating an external dictionary tool and training models end-to-end using reinforcement learning, in addition to supervised fine-tuning. Focusing on the Spanish-Wayuunaiki language pair, we frame translation as a tool-augmented decision-making problem in which the model can selectively consult a bilingual dictionary during generation. Our method combines supervised instruction tuning with Guided Reward Policy Optimization (GRPO), enabling the model to learn both when and how to use the tool effectively. BLEU similarity scores are used as rewards to guide this learning process. Preliminary results show that our tool-augmented models achieve up to +3.37 BLEU improvement over previous work, and a 18% relative gain compared to a supervised baseline without dictionary access, on the Spanish-Wayuunaiki test set from the AmericasNLP 2025 Shared Task. We also conduct ablation studies to assess the effects of model architecture and training strategy, comparing Qwen2.5-0.5B-Instruct with other models such as LLaMA and a prior NLLB-based system. These findings highlight the promise of combining LLMs with external tools and the role of reinforcement learning in improving translation quality in low-resource language settings.

35. Automatic Question & Answer Generation Using Generative Large Language Model (LLM)

Authors: Md. Alvee Ehsan, A.S.M Mehedi Hasan, Kefaya Benta Shahnoor, Syeda Sumaiya Tasneem
URL: https://arxiv.org/abs/2508.19475
Abstract:

\Abstract{In the realm of education, student evaluation holds equal significance as imparting knowledge. To be evaluated, students usually need to go through text-based academic assessment methods. Instructors need to make diverse sets of questions that need to be fair for all students to prove their adequacy over a particular topic. This can prove to be quite challenging as they may need to manually go through several different lecture materials. Our objective is to make this whole process much easier by implementing Automatic Question Answer Generation /(AQAG), using fine-tuned generative LLM. For tailoring the instructor’s preferred question style (MCQ, conceptual, or factual questions), prompt Engineering (PE) is being utilized. In this research, we propose to leverage unsupervised learning methods in NLP, primarily focusing on the English language. This approach empowers the base Meta-Llama 2-7B model to integrate RACE dataset as training data for the fine-tuning process. Creating a customized model that will offer efficient solutions for educators, instructors, and individuals engaged in text-based evaluations. A reliable and efficient tool for generating questions and answers can free up valuable time and resources, thus streamlining their evaluation processes.}

Authors: Sumon Kanti Dey, Jeanne M. Powell, Azra Ismail, Jeanmarie Perrone, Abeed Sarker
URL: https://arxiv.org/abs/2508.19467
Abstract:

Nonmedical opioid use is an urgent public health challenge, with far-reaching clinical and social consequences that are often underreported in traditional healthcare settings. Social media platforms, where individuals candidly share first-person experiences, offer a valuable yet underutilized source of insight into these impacts. In this study, we present a named entity recognition (NER) framework to extract two categories of self-reported consequences from social media narratives related to opioid use: ClinicalImpacts (e.g., withdrawal, depression) and SocialImpacts (e.g., job loss). To support this task, we introduce RedditImpacts 2.0, a high-quality dataset with refined annotation guidelines and a focus on first-person disclosures, addressing key limitations of prior work. We evaluate both fine-tuned encoder-based models and state-of-the-art large language models (LLMs) under zero- and few-shot in-context learning settings. Our fine-tuned DeBERTa-large model achieves a relaxed token-level F1 of 0.61 [95% CI: 0.43-0.62], consistently outperforming LLMs in precision, span accuracy, and adherence to task-specific guidelines. Furthermore, we show that strong NER performance can be achieved with substantially less labeled data, emphasizing the feasibility of deploying robust models in resource-limited settings. Our findings underscore the value of domain-specific fine-tuning for clinical NLP tasks and contribute to the responsible development of AI tools that may enhance addiction surveillance, improve interpretability, and support real-world healthcare decision-making. The best performing model, however, still significantly underperforms compared to inter-expert agreement (Cohen’s kappa: 0.81), demonstrating that a gap persists between expert intelligence and current state-of-the-art NER/AI capabilities for tasks requiring deep domain knowledge.

37. Bridging Language Gaps: Enhancing Few-Shot Language Adaptation

Authors: Philipp Borchert, Jochen De Weerdt, Marie-Francine Moens
URL: https://arxiv.org/abs/2508.19464
Abstract:

The disparity in language resources poses a challenge in multilingual NLP, with high-resource languages benefiting from extensive data, while low-resource languages lack sufficient data for effective training. Our Contrastive Language Alignment with Prompting (CoLAP) method addresses this gap by integrating contrastive learning with cross-lingual representations, facilitating task-specific knowledge transfer from high-resource to lower-resource languages. The primary advantage of our approach is its data efficiency, enabling rapid adaptation to new languages and reducing the need for large labeled datasets. We conduct experiments with multilingual encoder-only and decoder-only language models on natural language understanding tasks, including natural language inference and relation extraction, evaluating performance across both high- and low-resource languages. Our results demonstrate that CoLAP outperforms few-shot cross-lingual transfer baselines and in-context learning, even with limited available data. This effectively narrows the cross-lingual performance gap, contributing to the development of more efficient multilingual NLP techniques.

38. “She was useful, but a bit too optimistic”: Augmenting Design with Interactive Virtual Personas

Authors: Paluck Deep, Monica Bharadhidasan, A. Baki Kocaballi
URL: https://arxiv.org/abs/2508.19463
Abstract:

Personas have been widely used to understand and communicate user needs in human-centred design. Despite their utility, they may fail to meet the demands of iterative workflows due to their static nature, limited engagement, and inability to adapt to evolving design needs. Recent advances in large language models (LLMs) pave the way for more engaging and adaptive approaches to user representation. This paper introduces Interactive Virtual Personas (IVPs): multimodal, LLM-driven, conversational user simulations that designers can interview, brainstorm with, and gather feedback from in real time via voice interface. We conducted a qualitative study with eight professional UX designers, employing an IVP named “Alice” across three design activities: user research, ideation, and prototype evaluation. Our findings demonstrate the potential of IVPs to expedite information gathering, inspire design solutions, and provide rapid user-like feedback. However, designers raised concerns about biases, over-optimism, the challenge of ensuring authenticity without real stakeholder input, and the inability of the IVP to fully replicate the nuances of human interaction. Our participants emphasised that IVPs should be viewed as a complement to, not a replacement for, real user engagement. We discuss strategies for prompt engineering, human-in-the-loop integration, and ethical considerations for effective and responsible IVP use in design. Finally, our work contributes to the growing body of research on generative AI in the design process by providing insights into UX designers’ experiences of LLM-powered interactive personas.

39. A perishable ability? The future of writing in the face of generative artificial intelligence

Authors: Evandro L. T. P. Cunha
URL: https://arxiv.org/abs/2508.19427
Abstract:

The 2020s have been witnessing a very significant advance in the development of generative artificial intelligence tools, including text generation systems based on large language models. These tools have been increasingly used to generate texts in the most diverse domains – from technical texts to literary texts –, which might eventually lead to a lower volume of written text production by humans. This article discusses the possibility of a future in which human beings will have lost or significantly decreased their ability to write due to the outsourcing of this activity to machines. This possibility parallels the loss of the ability to write in other moments of human history, such as during the so-called Greek Dark Ages (approx. 1200 BCE - 800 BCE).

40. One Joke to Rule them All? On the (Im)possibility of Generalizing Humor

Authors: Mor Turgeman, Chen Shani, Dafna Shahaf
URL: https://arxiv.org/abs/2508.19402
Abstract:

Humor is a broad and complex form of communication that remains challenging for machines. Despite its broadness, most existing research on computational humor traditionally focused on modeling a specific type of humor. In this work, we wish to understand whether competence on one or more specific humor tasks confers any ability to transfer to novel, unseen types; in other words, is this fragmentation inevitable? This question is especially timely as new humor types continuously emerge in online and social media contexts (e.g., memes, anti-humor, AI fails). If Large Language Models (LLMs) are to keep up with this evolving landscape, they must be able to generalize across humor types by capturing deeper, transferable mechanisms. To investigate this, we conduct a series of transfer learning experiments across four datasets, representing different humor tasks. We train LLMs under varied diversity settings (1-3 datasets in training, testing on a novel task). Experiments reveal that models are capable of some transfer, and can reach up to 75% accuracy on unseen datasets; training on diverse sources improves transferability (1.88-4.05%) with minimal-to-no drop in in-domain performance. Further analysis suggests relations between humor types, with Dad Jokes surprisingly emerging as the best enabler of transfer (but is difficult to transfer to). We release data and code.

41. Fine-Tuning Vision-Language Models for Neutrino Event Analysis in High-Energy Physics Experiments

Authors: Dikshant Sagar, Kaiwen Yu, Alejandro Yankelevich, Jianming Bian, Pierre Baldi
URL: https://arxiv.org/abs/2508.19376
Abstract:

Recent progress in large language models (LLMs) has shown strong potential for multimodal reasoning beyond natural language. In this work, we explore the use of a fine-tuned Vision-Language Model (VLM), based on LLaMA 3.2, for classifying neutrino interactions from pixelated detector images in high-energy physics (HEP) experiments. We benchmark its performance against an established CNN baseline used in experiments like NOvA and DUNE, evaluating metrics such as classification accuracy, precision, recall, and AUC-ROC. Our results show that the VLM not only matches or exceeds CNN performance but also enables richer reasoning and better integration of auxiliary textual or semantic context. These findings suggest that VLMs offer a promising general-purpose backbone for event classification in HEP, paving the way for multimodal approaches in experimental neutrino physics.

42. Database Entity Recognition with Data Augmentation and Deep Learning

Authors: Zikun Fu, Chen Yang, Kourosh Davoudi, Ken Q. Pu
URL: https://arxiv.org/abs/2508.19372
Abstract:

This paper addresses the challenge of Database Entity Recognition (DB-ER) in Natural Language Queries (NLQ). We present several key contributions to advance this field: (1) a human-annotated benchmark for DB-ER task, derived from popular text-to-sql benchmarks, (2) a novel data augmentation procedure that leverages automatic annotation of NLQs based on the corresponding SQL queries which are available in popular text-to-SQL benchmarks, (3) a specialized language model based entity recognition model using T5 as a backbone and two down-stream DB-ER tasks: sequence tagging and token classification for fine-tuning of backend and performing DB-ER respectively. We compared our DB-ER tagger with two state-of-the-art NER taggers, and observed better performance in both precision and recall for our model. The ablation evaluation shows that data augmentation boosts precision and recall by over 10%, while fine-tuning of the T5 backbone boosts these metrics by 5-10%.

43. Grounding the Ungrounded: A Spectral-Graph Framework for Quantifying Hallucinations in multimodal LLMs

Authors: Supratik Sarkar, Swagatam Das
URL: https://arxiv.org/abs/2508.19366
Abstract:

Hallucinations in large language models (LLMs) remain a fundamental obstacle to trustworthy AI, particularly in high-stakes multimodal domains such as medicine, law, and finance. Existing evaluation techniques are largely heuristic – anchored in qualitative benchmarking or ad-hoc empirical mitigation – providing neither principled quantification nor actionable theoretical guarantees. This gap leaves a critical blind spot in understanding how hallucinations arise, propagate, and interact across modalities. We introduce the first (to our knowledge) rigorous information geometric framework in diffusion dynamics for quantifying hallucinations in multimodal LLMs (MLLMs), advancing the field from qualitative detection to mathematically grounded measurement. Our approach represents MLLM outputs as the spectral embeddings over multimodal graph Laplacians and characterizes the manifold gaps of truth vs inconsistencies as the semantic distortion, enabling the tight Rayleigh–Ritz bounds on the multimodal hallucination energy as a functional of time-dependent temperature profiles. By leveraging eigenmode decompositions in Reproducing Kernel Hilbert Space (RKHS) embeddings, our framework delivers modality-aware, theoretically interpretable metrics that capture the evolution of hallucinations across time and input prompts through temperature annealing. This work establishes a principled foundation for quantifying and bounding hallucinations, transforming them from a qualitative risk to a tractable, analyzable phenomenon.

44. LongReasonArena: A Long Reasoning Benchmark for Large Language Models

Authors: Jiayu Ding, Shuming Ma, Lei Cui, Nanning Zheng, Furu Wei
URL: https://arxiv.org/abs/2508.19363
Abstract:

Existing long-context benchmarks for Large Language Models (LLMs) focus on evaluating comprehension of long inputs, while overlooking the evaluation of long reasoning abilities. To address this gap, we introduce LongReasonArena, a benchmark specifically designed to assess the long reasoning capabilities of LLMs. Our tasks require models to solve problems by executing multi-step algorithms that reflect key aspects of long reasoning, such as retrieval and backtracking. By controlling the inputs, the required reasoning length can be arbitrarily scaled, reaching up to 1 million tokens of reasoning for the most challenging tasks. Extensive evaluation results demonstrate that LongReasonArena presents a significant challenge for both open-source and proprietary LLMs. For instance, Deepseek-R1 achieves only 7.5% accuracy on our task. Further analysis also reveals that the accuracy exhibits a linear decline with respect to the logarithm of the expected number of reasoning steps. Our code and data is available at this https URL.

45. Reflective Agreement: Combining Self-Mixture of Agents with a Sequence Tagger for Robust Event Extraction

Authors: Fatemeh Haji, Mazal Bethany, Cho-Yu Jason Chiang, Anthony Rios, Peyman Najafirad
URL: https://arxiv.org/abs/2508.19359
Abstract:

Event Extraction (EE) involves automatically identifying and extracting structured information about events from unstructured text, including triggers, event types, and arguments. Traditional discriminative models demonstrate high precision but often exhibit limited recall, particularly for nuanced or infrequent events. Conversely, generative approaches leveraging Large Language Models (LLMs) provide higher semantic flexibility and recall but suffer from hallucinations and inconsistent predictions. To address these challenges, we propose Agreement-based Reflective Inference System (ARIS), a hybrid approach combining a Self Mixture of Agents with a discriminative sequence tagger. ARIS explicitly leverages structured model consensus, confidence-based filtering, and an LLM reflective inference module to reliably resolve ambiguities and enhance overall event prediction quality. We further investigate decomposed instruction fine-tuning for enhanced LLM event extraction understanding. Experiments demonstrate our approach outperforms existing state-of-the-art event extraction methods across three benchmark datasets.

46. AT-CXR: Uncertainty-Aware Agentic Triage for Chest X-rays

Authors: Xueyang Li, Mingze Jiang, Gelei Xu, Jun Xia, Mengzhao Jia, Danny Chen, Yiyu Shi
URL: https://arxiv.org/abs/2508.19322
Abstract:

Agentic AI is advancing rapidly, yet truly autonomous medical-imaging triage, where a system decides when to stop, escalate, or defer under real constraints, remains relatively underexplored. To address this gap, we introduce AT-CXR, an uncertainty-aware agent for chest X-rays. The system estimates per-case confidence and distributional fit, then follows a stepwise policy to issue an automated decision or abstain with a suggested label for human intervention. We evaluate two router designs that share the same inputs and actions: a deterministic rule-based router and an LLM-decided router. Across five-fold evaluation on a balanced subset of NIH ChestX-ray14 dataset, both variants outperform strong zero-shot vision-language models and state-of-the-art supervised classifiers, achieving higher full-coverage accuracy and superior selective-prediction performance, evidenced by a lower area under the risk-coverage curve (AURC) and a lower error rate at high coverage, while operating with lower latency that meets practical clinical constraints. The two routers provide complementary operating points, enabling deployments to prioritize maximal throughput or maximal accuracy. Our code is available at this https URL.

47. An Investigation on Group Query Hallucination Attacks

Authors: Kehao Miao, Xiaolong Jin
URL: https://arxiv.org/abs/2508.19321
Abstract:

With the widespread use of large language models (LLMs), understanding their potential failure modes during user interactions is essential. In practice, users often pose multiple questions in a single conversation with LLMs. Therefore, in this study, we propose Group Query Attack, a technique that simulates this scenario by presenting groups of queries to LLMs simultaneously. We investigate how the accumulated context from consecutive prompts influences the outputs of LLMs. Specifically, we observe that Group Query Attack significantly degrades the performance of models fine-tuned on specific tasks. Moreover, we demonstrate that Group Query Attack induces a risk of triggering potential backdoors of LLMs. Besides, Group Query Attack is also effective in tasks involving reasoning, such as mathematical reasoning and code generation for pre-trained and aligned models.

48. MIDAS: Multimodal Interactive Digital-humAn Synthesis via Real-time Autoregressive Video Generation

Authors: Ming Chen, Liyuan Cui, Wenyuan Zhang, Haoxian Zhang, Yan Zhou, Xiaohan Li, Songlin Tang, Jiwen Liu, Borui Liao, Hejia Chen, Xiaoqiang Liu, Pengfei Wan
URL: https://arxiv.org/abs/2508.19320
Abstract:

Recently, interactive digital human video generation has attracted widespread attention and achieved remarkable progress. However, building such a practical system that can interact with diverse input signals in real time remains challenging to existing methods, which often struggle with heavy computational cost and limited controllability. In this work, we introduce an autoregressive video generation framework that enables interactive multimodal control and low-latency extrapolation in a streaming manner. With minimal modifications to a standard large language model (LLM), our framework accepts multimodal condition encodings including audio, pose, and text, and outputs spatially and semantically coherent representations to guide the denoising process of a diffusion head. To support this, we construct a large-scale dialogue dataset of approximately 20,000 hours from multiple sources, providing rich conversational scenarios for training. We further introduce a deep compression autoencoder with up to 64$\times$ reduction ratio, which effectively alleviates the long-horizon inference burden of the autoregressive model. Extensive experiments on duplex conversation, multilingual human synthesis, and interactive world model highlight the advantages of our approach in low latency, high efficiency, and fine-grained multimodal controllability.

49. DemoBias: An Empirical Study to Trace Demographic Biases in Vision Foundation Models

Authors: Abu Sufian, Anirudha Ghosh, Debaditya Barman, Marco Leo, Cosimo Distante
URL: https://arxiv.org/abs/2508.19298
Abstract:

Large Vision Language Models (LVLMs) have demonstrated remarkable capabilities across various downstream tasks, including biometric face recognition (FR) with description. However, demographic biases remain a critical concern in FR, as these foundation models often fail to perform equitably across diverse demographic groups, considering ethnicity/race, gender, and age. Therefore, through our work DemoBias, we conduct an empirical evaluation to investigate the extent of demographic biases in LVLMs for biometric FR with textual token generation tasks. We fine-tuned and evaluated three widely used pre-trained LVLMs: LLaVA, BLIP-2, and PaliGemma on our own generated demographic-balanced dataset. We utilize several evaluation metrics, like group-specific BERTScores and the Fairness Discrepancy Rate, to quantify and trace the performance disparities. The experimental results deliver compelling insights into the fairness and reliability of LVLMs across diverse demographic groups. Our empirical study uncovered demographic biases in LVLMs, with PaliGemma and LLaVA exhibiting higher disparities for Hispanic/Latino, Caucasian, and South Asian groups, whereas BLIP-2 demonstrated comparably consistent. Repository: this https URL.

50. Object Detection with Multimodal Large Vision-Language Models: An In-depth Review

Authors: Ranjan Sapkota, Manoj Karkee
URL: https://arxiv.org/abs/2508.19294
Abstract:

The fusion of language and vision in large vision-language models (LVLMs) has revolutionized deep learning-based object detection by enhancing adaptability, contextual reasoning, and generalization beyond traditional architectures. This in-depth review presents a structured exploration of the state-of-the-art in LVLMs, systematically organized through a three-step research review process. First, we discuss the functioning of vision language models (VLMs) for object detection, describing how these models harness natural language processing (NLP) and computer vision (CV) techniques to revolutionize object detection and localization. We then explain the architectural innovations, training paradigms, and output flexibility of recent LVLMs for object detection, highlighting how they achieve advanced contextual understanding for object detection. The review thoroughly examines the approaches used in integration of visual and textual information, demonstrating the progress made in object detection using VLMs that facilitate more sophisticated object detection and localization strategies. This review presents comprehensive visualizations demonstrating LVLMs’ effectiveness in diverse scenarios including localization and segmentation, and then compares their real-time performance, adaptability, and complexity to traditional deep learning systems. Based on the review, its is expected that LVLMs will soon meet or surpass the performance of conventional methods in object detection. The review also identifies a few major limitations of the current LVLM modes, proposes solutions to address those challenges, and presents a clear roadmap for the future advancement in this field. We conclude, based on this study, that the recent advancement in LVLMs have made and will continue to make a transformative impact on object detection and robotic applications in the future.

51. Stand on The Shoulders of Giants: Building JailExpert from Previous Attack Experience

Authors: Xi Wang, Songlei Jian, Shasha Li, Xiaopeng Li, Bin Ji, Jun Ma, Xiaodong Liu, Jing Wang, Feilong Bao, Jianfeng Zhang, Baosheng Wang, Jie Yu
URL: https://arxiv.org/abs/2508.19292
Abstract:

Large language models (LLMs) generate human-aligned content under certain safety constraints. However, the current known technique ``jailbreak prompt’’ can circumvent safety-aligned measures and induce LLMs to output malicious content. Research on Jailbreaking can help identify vulnerabilities in LLMs and guide the development of robust security frameworks. To circumvent the issue of attack templates becoming obsolete as models evolve, existing methods adopt iterative mutation and dynamic optimization to facilitate more automated jailbreak attacks. However, these methods face two challenges: inefficiency and repetitive optimization, as they overlook the value of past attack experiences. To better integrate past attack experiences to assist current jailbreak attempts, we propose the \textbf{JailExpert}, an automated jailbreak framework, which is the first to achieve a formal representation of experience structure, group experiences based on semantic drift, and support the dynamic updating of the experience pool. Extensive experiments demonstrate that JailExpert significantly improves both attack effectiveness and efficiency. Compared to the current state-of-the-art black-box jailbreak methods, JailExpert achieves an average increase of 17\% in attack success rate and 2.7 times improvement in attack efficiency. Our implementation is available at \href{this https URL}{XiZaiZai/JailExpert}

52. Seeing Like a Designer Without One: A Study on Unsupervised Slide Quality Assessment via Designer Cue Augmentation

Authors: Tai Inui, Steven Oh, Magdeline Kuan
URL: https://arxiv.org/abs/2508.19289
Abstract:

We present an unsupervised slide-quality assessment pipeline that combines seven expert-inspired visual-design metrics (whitespace, colorfulness, edge density, brightness contrast, text density, color harmony, layout balance) with CLIP-ViT embeddings, using Isolation Forest-based anomaly scoring to evaluate presentation slides. Trained on 12k professional lecture slides and evaluated on six academic talks (115 slides), our method achieved Pearson correlations up to 0.83 with human visual-quality ratings-1.79x to 3.23x stronger than scores from leading vision-language models (ChatGPT o4-mini-high, ChatGPT o3, Claude Sonnet 4, Gemini 2.5 Pro). We demonstrate convergent validity with visual ratings, discriminant validity against speaker-delivery scores, and exploratory alignment with overall impressions. Our results show that augmenting low-level design cues with multimodal embeddings closely approximates audience perceptions of slide quality, enabling scalable, objective feedback in real time.

53. Tricking LLM-Based NPCs into Spilling Secrets

Authors: Kyohei Shiomi, Zhuotao Lian, Toru Nakanishi, Teruaki Kitasuka
URL: https://arxiv.org/abs/2508.19288
Abstract:

Large Language Models (LLMs) are increasingly used to generate dynamic dialogue for game NPCs. However, their integration raises new security concerns. In this study, we examine whether adversarial prompt injection can cause LLM-based NPCs to reveal hidden background secrets that are meant to remain undisclosed.

54. Prompt-in-Content Attacks: Exploiting Uploaded Inputs to Hijack LLM Behavior

Authors: Zhuotao Lian, Weiyu Wang, Qingkui Zeng, Toru Nakanishi, Teruaki Kitasuka, Chunhua Su
URL: https://arxiv.org/abs/2508.19287
Abstract:

Large Language Models (LLMs) are widely deployed in applications that accept user-submitted content, such as uploaded documents or pasted text, for tasks like summarization and question answering. In this paper, we identify a new class of attacks, prompt in content injection, where adversarial instructions are embedded in seemingly benign inputs. When processed by the LLM, these hidden prompts can manipulate outputs without user awareness or system compromise, leading to biased summaries, fabricated claims, or misleading suggestions. We demonstrate the feasibility of such attacks across popular platforms, analyze their root causes including prompt concatenation and insufficient input isolation, and discuss mitigation strategies. Our findings reveal a subtle yet practical threat in real-world LLM workflows.

55. RL-Finetuned LLMs for Privacy-Preserving Synthetic Rewriting

Authors: Zhan Shi, Yefeng Yuan, Yuhong Liu, Liang Cheng, Yi Fang
URL: https://arxiv.org/abs/2508.19286
Abstract:

The performance of modern machine learning systems depends on access to large, high-quality datasets, often sourced from user-generated content or proprietary, domain-specific corpora. However, these rich datasets inherently contain sensitive personal information, raising significant concerns about privacy, data security, and compliance with regulatory frameworks. While conventional anonymization techniques can remove explicit identifiers, such removal may result in performance drop in downstream machine learning tasks. More importantly, simple anonymization may not be effective against inference attacks that exploit implicit signals such as writing style, topical focus, or demographic cues, highlighting the need for more robust privacy safeguards during model training. To address the challenging issue of balancing user privacy and data utility, we propose a reinforcement learning framework that fine-tunes a large language model (LLM) using a composite reward function that jointly optimizes for explicit and implicit privacy, semantic fidelity, and output diversity. To effectively capture population level regularities, the privacy reward combines semantic cues with structural patterns derived from a minimum spanning tree (MST) over latent representations. By modeling these privacy-sensitive signals in their distributional context, the proposed approach guides the model to generate synthetic rewrites that preserve utility while mitigating privacy risks. Empirical results show that the proposed method significantly enhances author obfuscation and privacy metrics without degrading semantic quality, providing a scalable and model-agnostic solution for privacy preserving data generation in the era of large language models.

56. CORE: Lossless Compression for Retrieval-Augmented LLMs via Reinforcement Learning

Authors: Ziqiang Cui, Yunpeng Weng, Xing Tang, Peiyang Liu, Shiwei Li, Bowei He, Jiamin Chen, Xiuqiang He, Chen Ma
URL: https://arxiv.org/abs/2508.19282
Abstract:

Retrieval-Augmented Generation (RAG) has emerged as a promising approach to enhance the timeliness of knowledge and the factual accuracy of responses in Large Language Models (LLMs). However, the inclusion of excessive retrieved documents substantially increases the input length, leading to higher computational costs. Previous studies have attempted to compress retrieved documents into shorter texts before in-context integration, but such methods often compromise end-task performance. The lack of well-defined compression targets forces many approaches to rely on fixed heuristics, which cannot guarantee that the compressed content will effectively support the end task. To address these limitations, we propose CORE, a novel method designed to achieve lossless context compression for RAG. CORE employs reinforcement learning to optimize the compression process without relying on predefined compression labels. Specifically, it utilizes end-task performance as a reward signal and applies Generalized Reinforcement Learning Policy Optimization (GRPO) to train the compressor. This end-to-end training framework enables the compressor to generate summaries that maximize the accuracy of answers generated by the LLM. Extensive experiments on four datasets demonstrate the superiority of our approach. With a high compression ratio of 3\%, our method not only avoids performance degradation compared to prepending full documents across all datasets but also improves the average Exact Match (EM) score by 3.3 points. The code will be released soon.

Authors: Gunjan Jalori, Preetika Verma, Sercan Ö Arık
URL: https://arxiv.org/abs/2508.19279
Abstract:

Time series Forecasting with large languagemodels (LLMs) requires bridging numericalpatterns and natural language. Effective fore-casting on LLM often relies on extensive pre-processing and this http URL studiesshow that a frozen LLM can rival specializedforecasters when supplied with a carefully en-gineered natural-language prompt, but craft-ing such a prompt for each task is itself oner-ous and ad-hoc. We introduce FLAIRR-TS, atest-time prompt optimization framework thatutilizes an agentic system: a Forecaster-agentgenerates forecasts using an initial prompt,which is then refined by a refiner agent, in-formed by past outputs and retrieved this http URL adaptive prompting generalizes across do-mains using creative prompt templates andgenerates high-quality forecasts without inter-mediate code this http URL onbenchmark datasets show improved accuracyover static prompting and retrieval-augmentedbaselines, approaching the performance ofspecialized this http URL-TS providesa practical alternative to tuning, achievingstrong performance via its agentic approach toadaptive prompt refinement and retrieval.

58. POT: Inducing Overthinking in LLMs via Black-Box Iterative Optimization

Authors: Xinyu Li, Tianjin Huang, Ronghui Mu, Xiaowei Huang, Gaojie Jin
URL: https://arxiv.org/abs/2508.19277
Abstract:

Recent advances in Chain-of-Thought (CoT) prompting have substantially enhanced the reasoning capabilities of large language models (LLMs), enabling sophisticated problem-solving through explicit multi-step reasoning traces. However, these enhanced reasoning processes introduce novel attack surfaces, particularly vulnerabilities to computational inefficiency through unnecessarily verbose reasoning chains that consume excessive resources without corresponding performance gains. Prior overthinking attacks typically require restrictive conditions including access to external knowledge sources for data poisoning, reliance on retrievable poisoned content, and structurally obvious templates that limit practical applicability in real-world scenarios. To address these limitations, we propose POT (Prompt-Only OverThinking), a novel black-box attack framework that employs LLM-based iterative optimization to generate covert and semantically natural adversarial prompts, eliminating dependence on external data access and model retrieval. Extensive experiments across diverse model architectures and datasets demonstrate that POT achieves superior performance compared to other methods.

59. Rethinking Reasoning in LLMs: Neuro-Symbolic Local RetoMaton Beyond ICL and CoT

Authors: Rushitha Santhoshi Mamidala, Anshuman Chhabra, Ankur Mali
URL: https://arxiv.org/abs/2508.19271
Abstract:

Prompt-based reasoning strategies such as Chain-of-Thought (CoT) and In-Context Learning (ICL) have become widely used for eliciting reasoning capabilities in large language models (LLMs). However, these methods rely on fragile, implicit mechanisms often yielding inconsistent outputs across seeds, formats, or minor prompt variations making them fundamentally unreliable for tasks requiring stable, interpretable reasoning. In contrast, automata-based neuro-symbolic frameworks like RetoMaton offer a more structured and trustworthy alternative by grounding retrieval in symbolic memory with deterministic transitions. In this work, we extend RetoMaton by replacing its global datastore with a local, task-adaptive Weighted Finite Automaton (WFA), constructed directly from external domain corpora. This local automaton structure promotes robust, context-aware retrieval while preserving symbolic traceability and low inference overhead. Unlike prompting, which entangles context and memory in opaque ways, our approach leverages the explicit structure of WFAs to provide verifiable and modular retrieval behavior, making it better suited for domain transfer and interoperability. We evaluate this local RetoMaton variant on two pretrained LLMs LLaMA-3.2-1B and Gemma-3-1B-PT across three reasoning tasks: TriviaQA (reading comprehension), GSM8K (multi-step math), and MMLU (domain knowledge). Compared to the base model and prompting-based methods, augmenting these setups with local RetoMaton consistently improves performance while enabling transparent and reproducible retrieval dynamics. Our results highlight a promising shift toward trustworthy, symbolic reasoning in modern LLMs via lightweight, automaton-guided memory.

60. Should LLMs be WEIRD? Exploring WEIRDness and Human Rights in Large Language Models

Authors: Ke Zhou, Marios Constantinides, Daniele Quercia
URL: https://arxiv.org/abs/2508.19269
Abstract:

Large language models (LLMs) are often trained on data that reflect WEIRD values: Western, Educated, Industrialized, Rich, and Democratic. This raises concerns about cultural bias and fairness. Using responses to the World Values Survey, we evaluated five widely used LLMs: GPT-3.5, GPT-4, Llama-3, BLOOM, and Qwen. We measured how closely these responses aligned with the values of the WEIRD countries and whether they conflicted with human rights principles. To reflect global diversity, we compared the results with the Universal Declaration of Human Rights and three regional charters from Asia, the Middle East, and Africa. Models with lower alignment to WEIRD values, such as BLOOM and Qwen, produced more culturally varied responses but were 2% to 4% more likely to generate outputs that violated human rights, especially regarding gender and equality. For example, some models agreed with the statements a man who cannot father children is not a real man'' anda husband should always know where his wife is’’, reflecting harmful gender norms. These findings suggest that as cultural representation in LLMs increases, so does the risk of reproducing discriminatory beliefs. Approaches such as Constitutional AI, which could embed human rights principles into model behavior, may only partly help resolve this tension.

61. MultiPL-MoE: Multi-Programming-Lingual Extension of Large Language Models through Hybrid Mixture-of-Experts

Authors: Qing Wang, Xue Han, Jiahui Wang, Lehao Xing, Qian Hu, Lianlian Zhang, Chao Deng, Junlan Feng
URL: https://arxiv.org/abs/2508.19268
Abstract:

Despite LLMs’ excellent code creation capabilities, multilingual code generation remains extremely challenging. To address this, we intent to improve the multi-programming-lingual (MultiPL) performance of the base LLMs while retaining the most popular ones using restricted computational resources. We consider MultiPL to be a special case of multiple natural languages and propose a MultiPL extension of LLMs utilizing a hybrid mixture of experts (MoE), called MultiPL-MoE. Specifically, MultiPL-MoE combines two paired MoEs to optimize expert selection at both the token and segment levels. The token-level MoE is a standard upcycling MoE structure with a shared expert and a novel gate weight normalization approach that aids in the final fusion with the segment-level MoE. The segment-level MoE incorporates two innovative designs to better capture the syntactic structure and contextual patterns of programming languages: First, using a sliding window to partition the input token sequence into multiple segments; Then, adopting an expert-choice routing strategy that allows experts to select the top-k segments. The results of the experiment proved the effectiveness of MultiPL-MoE.

62. Lossless Compression of Neural Network Components: Weights, Checkpoints, and K/V Caches in Low-Precision Formats

Authors: Anat Heilper, Doron Singer
URL: https://arxiv.org/abs/2508.19263
Abstract:

As deep learning models grow and deployment becomes more widespread, reducing the storage and transmission costs of neural network weights has become increasingly important. While prior work such as ZipNN has shown that lossless compression methods - particularly those based on Huffman encoding floating-point exponents can significantly reduce model sizes, these techniques have primarily been applied to higher-precision formats such as FP32 and BF16. In this work, we extend the ZipNN approach to lower-precision floating-point formats, specifically FP8 and FP4, which are gaining popularity for efficient inference. We design a compression method that separates and compresses the exponent and mantissa components independently using entropy coding. Our evaluation shows compression ratios up to 62% for BF16 and 83% for FP8. We also investigate the compressibility of key-value (K/V) cache tensors used in large language models (LLMs), finding that they, too, exhibit compressible patterns, enabling memory savings during deployment.

63. Real-Time Intuitive AI Drawing System for Collaboration: Enhancing Human Creativity through Formal and Contextual Intent Integration

Authors: Jookyung Song, Mookyoung Kang, Nojun Kwak
URL: https://arxiv.org/abs/2508.19254
Abstract:

This paper presents a real-time generative drawing system that interprets and integrates both formal intent - the structural, compositional, and stylistic attributes of a sketch - and contextual intent - the semantic and thematic meaning inferred from its visual content - into a unified transformation process. Unlike conventional text-prompt-based generative systems, which primarily capture high-level contextual descriptions, our approach simultaneously analyzes ground-level intuitive geometric features such as line trajectories, proportions, and spatial arrangement, and high-level semantic cues extracted via vision-language models. These dual intent signals are jointly conditioned in a multi-stage generation pipeline that combines contour-preserving structural control with style- and content-aware image synthesis. Implemented with a touchscreen-based interface and distributed inference architecture, the system achieves low-latency, two-stage transformation while supporting multi-user collaboration on shared canvases. The resulting platform enables participants, regardless of artistic expertise, to engage in synchronous, co-authored visual creation, redefining human-AI interaction as a process of co-creation and mutual enhancement.

64. Federated Fine-Tuning of Sparsely-Activated Large Language Models on Resource-Constrained Devices

Authors: Fahao Chen, Jie Wan, Peng Li, Zhou Su, Dongxiao Yu
URL: https://arxiv.org/abs/2508.19078
Abstract:

Federated fine-tuning of Mixture-of-Experts (MoE)-based large language models (LLMs) is challenging due to their massive computational requirements and the resource constraints of participants. Existing working attempts to fill this gap through model quantization, computation offloading, or expert pruning. However, they cannot achieve desired performance due to impractical system assumptions and a lack of consideration for MoE-specific characteristics. In this paper, we propose FLUX, a system designed to enable federated fine-tuning of MoE-based LLMs across participants with constrained computing resources (e.g., consumer-grade GPUs), aiming to minimize time-to-accuracy. FLUX introduces three key innovations: (1) quantization-based local profiling to estimate expert activation with minimal overhead, (2) adaptive layer-aware expert merging to reduce resource consumption while preserving accuracy, and (3) dynamic expert role assignment using an exploration-exploitation strategy to balance tuning and non-tuning experts. Extensive experiments on LLaMA-MoE and DeepSeek-MoE with multiple benchmark datasets demonstrate that FLUX significantly outperforms existing methods, achieving up to 4.75X speedup in time-to-accuracy.

65. MovieCORE: COgnitive REasoning in Movies

Authors: Gueter Josmy Faure, Min-Hung Chen, Jia-Fong Yeh, Ying Cheng, Hung-Ting Su, Yung-Hao Tang, Shang-Hong Lai, Winston H. Hsu
URL: https://arxiv.org/abs/2508.19026
Abstract:

This paper introduces MovieCORE, a novel video question answering (VQA) dataset designed to probe deeper cognitive understanding of movie content. Unlike existing datasets that focus on surface-level comprehension, MovieCORE emphasizes questions that engage System-2 thinking while remaining specific to the video material. We present an innovative agentic brainstorming approach, utilizing multiple large language models (LLMs) as thought agents to generate and refine high-quality question-answer pairs. To evaluate dataset quality, we develop a set of cognitive tests assessing depth, thought-provocation potential, and syntactic complexity. We also propose a comprehensive evaluation scheme for assessing VQA model performance on deeper cognitive tasks. To address the limitations of existing video-language models (VLMs), we introduce an agentic enhancement module, Agentic Choice Enhancement (ACE), which improves model reasoning capabilities post-training by up to 25%. Our work contributes to advancing movie understanding in AI systems and provides valuable insights into the capabilities and limitations of current VQA models when faced with more challenging, nuanced questions about cinematic content. Our project page, dataset and code can be found at this https URL.

LLM 관련 주요 논문 - 2025-08-29

1. SWIRL: A Staged Workflow for Interleaved Reinforcement Learning in Mobile GUI Control

2. Tracking World States with Language Models: State-Based Evaluation Using Chess

3. InquireMobile: Teaching VLM-based Mobile Agent to Request Human Assistance via Reinforcement Fine-Tuning

4. Instructional Agents: LLM Agents on Automated Course Material Generation for Teaching Faculties

5. ReST-RL: Achieving Accurate Code Reasoning of LLMs with Optimized Self-Training and Decoding

6. Democracy-in-Silico: Institutional Design as Alignment in AI-Governed Polities

7. Caught in the Act: a mechanistic approach to detecting deception

8. SLIM: Subtrajectory-Level Elimination for More Effective Reasoning

9. Reliable Weak-to-Strong Monitoring of LLM Agents

10. Quantized but Deceptive? A Multi-Dimensional Truthfulness Evaluation of Quantized LLMs

11. Large Language Models (LLMs) for Electronic Design Automation (EDA)

12. Symphony: A Decentralized Multi-Agent Framework for Scalable Collective Intelligence

13. Decomposing Behavioral Phase Transitions in LLMs: Order Parameters for Emergent Misalignment

14. MathBuddy: A Multimodal System for Affective Math Tutoring

15. Diffusion Language Models Know the Answer Before Decoding

16. GLSim: Detecting Object Hallucinations in LVLMs via Global-Local Similarity

17. Dhati+: Fine-tuned Large Language Models for Arabic Subjectivity Evaluation

18. Logical Reasoning with Outcome Reward Models for Test-Time Scaling

19. AI-Powered Detection of Inappropriate Language in Medical School Curricula

20. SoK: Large Language Model Copyright Auditing via Fingerprinting

21. PSO-Merging: Merging Models Based on Particle Swarm Optimization

22. Bootstrapping Learned Cost Models with Synthetic SQL Queries

23. NLKI: A lightweight Natural Language Knowledge Integration Framework for Improving Small VLMs in Commonsense VQA Tasks

24. Safety Alignment Should Be Made More Than Just A Few Attention Heads

25. Survey of Specialized Large Language Model

26. LFD: Layer Fused Decoding to Exploit External Knowledge in Retrieval-Augmented Generation

27. Towards a Holistic and Automated Evaluation Framework for Multi-Level Comprehension of LLMs in Book-Length Contexts

28. Generative Models for Synthetic Data: Transforming Data Mining in the GenAI Era

29. Just Because You Can, Doesn’t Mean You Should: LLMs for Data Fitting

30. Taming the Chaos: Coordinated Autoscaling for Heterogeneous and Disaggregated LLM Inference

31. Language Models Identify Ambiguities and Exploit Loopholes

32. Orchid: Orchestrating Context Across Creative Workflows with Generative AI

33. Learning Game-Playing Agents with Generative Code Optimization

34. Improving Low-Resource Translation with Dictionary-Guided Fine-Tuning and RL: A Spanish-to-Wayuunaiki Study

35. Automatic Question & Answer Generation Using Generative Large Language Model (LLM)

36. Inference Gap in Domain Expertise and Machine Intelligence in Named Entity Recognition: Creation of and Insights from a Substance Use-related Dataset

37. Bridging Language Gaps: Enhancing Few-Shot Language Adaptation

38. “She was useful, but a bit too optimistic”: Augmenting Design with Interactive Virtual Personas

39. A perishable ability? The future of writing in the face of generative artificial intelligence

40. One Joke to Rule them All? On the (Im)possibility of Generalizing Humor

41. Fine-Tuning Vision-Language Models for Neutrino Event Analysis in High-Energy Physics Experiments

42. Database Entity Recognition with Data Augmentation and Deep Learning

43. Grounding the Ungrounded: A Spectral-Graph Framework for Quantifying Hallucinations in multimodal LLMs

44. LongReasonArena: A Long Reasoning Benchmark for Large Language Models

45. Reflective Agreement: Combining Self-Mixture of Agents with a Sequence Tagger for Robust Event Extraction

46. AT-CXR: Uncertainty-Aware Agentic Triage for Chest X-rays

47. An Investigation on Group Query Hallucination Attacks

48. MIDAS: Multimodal Interactive Digital-humAn Synthesis via Real-time Autoregressive Video Generation

49. DemoBias: An Empirical Study to Trace Demographic Biases in Vision Foundation Models

50. Object Detection with Multimodal Large Vision-Language Models: An In-depth Review

51. Stand on The Shoulders of Giants: Building JailExpert from Previous Attack Experience

52. Seeing Like a Designer Without One: A Study on Unsupervised Slide Quality Assessment via Designer Cue Augmentation

53. Tricking LLM-Based NPCs into Spilling Secrets

54. Prompt-in-Content Attacks: Exploiting Uploaded Inputs to Hijack LLM Behavior

55. RL-Finetuned LLMs for Privacy-Preserving Synthetic Rewriting

56. CORE: Lossless Compression for Retrieval-Augmented LLMs via Reinforcement Learning

57. FLAIRR-TS – Forecasting LLM-Agents with Iterative Refinement and Retrieval for Time Series

58. POT: Inducing Overthinking in LLMs via Black-Box Iterative Optimization

59. Rethinking Reasoning in LLMs: Neuro-Symbolic Local RetoMaton Beyond ICL and CoT

60. Should LLMs be WEIRD? Exploring WEIRDness and Human Rights in Large Language Models

61. MultiPL-MoE: Multi-Programming-Lingual Extension of Large Language Models through Hybrid Mixture-of-Experts

62. Lossless Compression of Neural Network Components: Weights, Checkpoints, and K/V Caches in Low-Precision Formats

63. Real-Time Intuitive AI Drawing System for Collaboration: Enhancing Human Creativity through Formal and Contextual Intent Integration

64. Federated Fine-Tuning of Sparsely-Activated Large Language Models on Resource-Constrained Devices

65. MovieCORE: COgnitive REasoning in Movies