Authors: Luca Fantin, Marco Antonelli, Margherita Cesetti, Daniele Irto, Bruno Zamengo, Francesco Silvestri
Abstract: Assessing the quality of public transportation services requires the analysis of large quantities of data on the scheduled and actual trips and documents listing the quality constraints each service needs to meet. Interrogating such datasets with SQL queries, organizing and visualizing the data can be quite complex for most users. This paper presents a chatbot offering a user-friendly tool to interact with these datasets and support decision making. It is based on an agent architecture, which expands the capabilities of the core Large Language Model (LLM) by allowing it to interact with a series of tools that can execute several tasks, like performing SQL queries, plotting data and creating maps from the coordinates of a trip and its stops. This paper also tackles one of the main open problems of such Generative AI projects: collecting data to measure the system's performance. Our chatbot has been extensively tested with a workflow that asks several questions and stores the generated query, the retrieved data and the natural language response for each of them. Such questions are drawn from a set of base examples which are then completed with actual data from the database. This procedure yields a dataset for the evaluation of the chatbot's performance, especially the consistency of its answers and the correctness of the generated queries.
Authors: Arseniy Pertzovsky, Roni Stern, Ariel Felner, Roie Zivan
Abstract: We explore the use of Artificial Potential Fields (APFs) to solve Multi-Agent Path Finding (MAPF) and Lifelong MAPF (LMAPF) problems. In MAPF, a team of agents must move to their goal locations without collisions, whereas in LMAPF, new goals are generated upon arrival. We propose methods for incorporating APFs in a range of MAPF algorithms, including Prioritized Planning, MAPF-LNS2, and Priority Inheritance with Backtracking (PIBT). Experimental results show that using APF is not beneficial for MAPF but yields up to a 7-fold increase in overall system throughput for LMAPF.
Authors: Tian Qin, Core Francisco Park, Mujin Kwun, Aaron Walsman, Eran Malach, Nikhil Anand, Hidenori Tanaka, David Alvarez-Melis
Abstract: Mathematical reasoning tasks have become prominent benchmarks for assessing the reasoning capabilities of LLMs, especially with reinforcement learning (RL) methods such as GRPO showing significant performance gains. However, accuracy metrics alone do not support fine-grained assessment of capabilities and fail to reveal which problem-solving skills have been internalized. To better understand these capabilities, we propose to decompose problem solving into fundamental capabilities: Plan (mapping questions to sequences of steps), Execute (correctly performing solution steps), and Verify (identifying the correctness of a solution). Empirically, we find that GRPO mainly enhances the execution skill-improving execution robustness on problems the model already knows how to solve-a phenomenon we call temperature distillation. More importantly, we show that RL-trained models struggle with fundamentally new problems, hitting a 'coverage wall' due to insufficient planning skills. To explore RL's impact more deeply, we construct a minimal, synthetic solution-tree navigation task as an analogy for mathematical problem-solving. This controlled setup replicates our empirical findings, confirming RL primarily boosts execution robustness. Importantly, in this setting, we identify conditions under which RL can potentially overcome the coverage wall through improved exploration and generalization to new solution paths. Our findings provide insights into the role of RL in enhancing LLM reasoning, expose key limitations, and suggest a path toward overcoming these barriers. Code is available at https://github.com/cfpark00/RL-Wall.
Authors: Mohammad Helal Uddin, Sabur Baidya
Abstract: Mental disorders including depression, anxiety, and other neurological disorders pose a significant global challenge, particularly among individuals exhibiting social avoidance tendencies. This study proposes a hybrid approach by leveraging smartphone sensor data measuring daily physical activities and analyzing their social media (Twitter) interactions for evaluating an individual's depression level. Using CNN-based deep learning models and Naive Bayes classification, we identify human physical activities accurately and also classify the user sentiments. A total of 33 participants were recruited for data acquisition, and nine relevant features were extracted from the physical activities and analyzed with their weekly depression scores, evaluated using the Geriatric Depression Scale (GDS) questionnaire. Of the nine features, six are derived from physical activities, achieving an activity recognition accuracy of 95%, while three features stem from sentiment analysis of Twitter activities, yielding a sentiment analysis accuracy of 95.6%. Notably, several physical activity features exhibited significant correlations with the severity of depression symptoms. For classifying the depression severity, a support vector machine (SVM)-based algorithm is employed that demonstrated a very high accuracy of 94%, outperforming alternative models, e.g., the multilayer perceptron (MLP) and k-nearest neighbor. It is a simple approach yet highly effective in the long run for monitoring depression without breaching personal privacy.
Authors: Yuval David, Fabiana Fournier, Lior Limonad, Inna Skarbovsky
Abstract: Causal reasoning is essential for business process interventions and improvement, requiring a clear understanding of causal relationships among activity execution times in an event log. Recent work introduced a method for discovering causal process models but lacked the ability to capture alternating causal conditions across multiple variants. This raises the challenges of handling missing values and expressing the alternating conditions among log splits when blending traces with varying activities. We propose a novel method to unify multiple causal process variants into a consistent model that preserves the correctness of the original causal models, while explicitly representing their causal-flow alternations. The method is formally defined, proved, evaluated on three open and two proprietary datasets, and released as an open-source implementation.
Authors: Massimiliano Pronesti, Michela Lorandi, Paul Flanagan, Oisin Redmon, Anya Belz, Yufang Hou
Abstract: Systematic reviews in medicine play a critical role in evidence-based decision-making by aggregating findings from multiple studies. A central bottleneck in automating this process is extracting numeric evidence and determining study-level conclusions for specific outcomes and comparisons. Prior work has framed this problem as a textual inference task by retrieving relevant content fragments and inferring conclusions from them. However, such approaches often rely on shallow textual cues and fail to capture the underlying numeric reasoning behind expert assessments. In this work, we conceptualise the problem as one of quantitative reasoning. Rather than inferring conclusions from surface text, we extract structured numerical evidence (e.g., event counts or standard deviations) and apply domain knowledge informed logic to derive outcome-specific conclusions. We develop a numeric reasoning system composed of a numeric data extraction model and an effect estimate component, enabling more accurate and interpretable inference aligned with the domain expert principles. We train the numeric data extraction model using different strategies, including supervised fine-tuning (SFT) and reinforcement learning (RL) with a new value reward model. When evaluated on the CochraneForest benchmark, our best-performing approach -- using RL to train a small-scale number extraction model -- yields up to a 21% absolute improvement in F1 score over retrieval-based systems and outperforms general-purpose LLMs of over 400B parameters by up to 9%. Our results demonstrate the promise of reasoning-driven approaches for automating systematic evidence synthesis.
Authors: Michael Sun, Weize Yuan, Gang Liu, Wojciech Matusik, Jie Chen
Abstract: Recent data-efficient molecular generation approaches exploit graph grammars to introduce interpretability into the generative models. However, grammar learning therein relies on expert annotation or unreliable heuristics for algorithmic inference. We propose Foundation Molecular Grammar (FMG), which leverages multi-modal foundation models (MMFMs) to induce an interpretable molecular language. By exploiting the chemical knowledge of an MMFM, FMG renders molecules as images, describes them as text, and aligns information across modalities using prompt learning. FMG can be used as a drop-in replacement for the prior grammar learning approaches in molecular generation and property prediction. We show that FMG not only excels in synthesizability, diversity, and data efficiency but also offers built-in chemical interpretability for automated molecular discovery workflows. Code is available at https://github.com/shiningsunnyday/induction.
Authors: Jenny Zhang, Shengran Hu, Cong Lu, Robert Lange, Jeff Clune
Abstract: Today's AI systems have human-designed, fixed architectures and cannot autonomously and continuously improve themselves. The advance of AI could itself be automated. If done safely, that would accelerate AI development and allow us to reap its benefits much sooner. Meta-learning can automate the discovery of novel algorithms, but is limited by first-order improvements and the human design of a suitable search space. The G\"odel machine proposed a theoretical alternative: a self-improving AI that repeatedly modifies itself in a provably beneficial manner. Unfortunately, proving that most changes are net beneficial is impossible in practice. We introduce the Darwin G\"odel Machine (DGM), a self-improving system that iteratively modifies its own code (thereby also improving its ability to modify its own codebase) and empirically validates each change using coding benchmarks. Inspired by Darwinian evolution and open-endedness research, the DGM maintains an archive of generated coding agents. It grows the archive by sampling an agent from it and using a foundation model to create a new, interesting, version of the sampled agent. This open-ended exploration forms a growing tree of diverse, high-quality agents and allows the parallel exploration of many different paths through the search space. Empirically, the DGM automatically improves its coding capabilities (e.g., better code editing tools, long-context window management, peer-review mechanisms), increasing performance on SWE-bench from 20.0% to 50.0%, and on Polyglot from 14.2% to 30.7%. Furthermore, the DGM significantly outperforms baselines without self-improvement or open-ended exploration. All experiments were done with safety precautions (e.g., sandboxing, human oversight). The DGM is a significant step toward self-improving AI, capable of gathering its own stepping stones along paths that unfold into endless innovation.
Authors: Yongjin Yang, Euiin Yi, Jongwoo Ko, Kimin Lee, Zhijing Jin, Se-Young Yun
Abstract: The remarkable growth in large language model (LLM) capabilities has spurred exploration into multi-agent systems, with debate frameworks emerging as a promising avenue for enhanced problem-solving. These multi-agent debate (MAD) approaches, where agents collaboratively present, critique, and refine arguments, potentially offer improved reasoning, robustness, and diverse perspectives over monolithic models. Despite prior studies leveraging MAD, a systematic understanding of its effectiveness compared to self-agent methods, particularly under varying conditions, remains elusive. This paper seeks to fill this gap by conceptualizing MAD as a test-time computational scaling technique, distinguished by collaborative refinement and diverse exploration capabilities. We conduct a comprehensive empirical investigation comparing MAD with strong self-agent test-time scaling baselines on mathematical reasoning and safety-related tasks. Our study systematically examines the influence of task difficulty, model scale, and agent diversity on MAD's performance. Key findings reveal that, for mathematical reasoning, MAD offers limited advantages over self-agent scaling but becomes more effective with increased problem difficulty and decreased model capability, while agent diversity shows little benefit. Conversely, for safety tasks, MAD's collaborative refinement can increase vulnerability, but incorporating diverse agent configurations facilitates a gradual reduction in attack success through the collaborative refinement process. We believe our findings provide critical guidance for the future development of more effective and strategically deployed MAD systems.
Authors: Nick Byrd
Abstract: By late 20th century, the rationality wars had launched debates about the nature and norms of intuitive and reflective thinking. Those debates drew from mid-20th century ideas such as bounded rationality, which challenged more idealized notions of rationality observed since the 19th century. Now that 21st century cognitive scientists are applying the resulting dual process theories to artificial intelligence, it is time to dust off some lessons from this history. So this paper synthesizes old ideas with recent results from experiments on humans and machines. The result is Strategic Reflectivism, which takes the position that one key to intelligent systems (human or artificial) is pragmatic switching between intuitive and reflective inference to optimally fulfill competing goals. Strategic Reflectivism builds on American Pragmatism, transcends superficial indicators of reflective thinking such as model size or chains of thought, and becomes increasingly actionable as we learn more about the value of intuition and reflection.
Authors: Pin-Han Chen, Yu-Sheng Lin, Wei-Cheng Lee, Tin-Yu Leu, Po-Hsiang Hsu, Anjana Dissanayake, Sungjin Oh, Chinq-Shiun Chiu
Abstract: RF/Analog design is essential for bridging digital technologies with real-world signals, ensuring the functionality and reliability of a wide range of electronic systems. However, analog design procedures are often intricate, time-consuming and reliant on expert intuition, and hinder the time and cost efficiency of circuit development. To overcome the limitations of the manual circuit design, we introduce MenTeR - a multiagent workflow integrated into an end-to-end analog design framework. By employing multiple specialized AI agents that collaboratively address different aspects of the design process, such as specification understanding, circuit optimization, and test bench validation, MenTeR reduces the dependency on frequent trial-and-error-style intervention. MenTeR not only accelerates the design cycle time but also facilitates a broader exploration of the design space, demonstrating robust capabilities in handling real-world analog systems. We believe that MenTeR lays the groundwork for future "RF/Analog Copilots" that can collaborate seamlessly with human designers.
Authors: Guangyi Liu, Yongqi Zhang, Xunyuan Liu, Quanming Yao
Abstract: Drug-drug interaction (DDI) prediction is critical for treatment safety. While large language models (LLMs) show promise in pharmaceutical tasks, their effectiveness in DDI prediction remains challenging. Inspired by the well-established clinical practice where physicians routinely reference similar historical cases to guide their decisions through case-based reasoning (CBR), we propose CBR-DDI, a novel framework that distills pharmacological principles from historical cases to improve LLM reasoning for DDI tasks. CBR-DDI constructs a knowledge repository by leveraging LLMs to extract pharmacological insights and graph neural networks (GNNs) to model drug associations. A hybrid retrieval mechanism and dual-layer knowledge-enhanced prompting allow LLMs to effectively retrieve and reuse relevant cases. We further introduce a representative sampling strategy for dynamic case refinement. Extensive experiments demonstrate that CBR-DDI achieves state-of-the-art performance, with a significant 28.7% accuracy improvement over both popular LLMs and CBR baseline, while maintaining high interpretability and flexibility.
Authors: Yutong Xie, Zhuoheng Li, Xiyuan Wang, Yijun Pan, Qijia Liu, Xingzhi Cui, Kuang-Yu Lo, Ruoyi Gao, Xingjian Zhang, Jin Huang, Walter Yuan, Matthew O. Jackson, Qiaozhu Mei
Abstract: Despite their success in numerous fields, the potential of foundation models for modeling and understanding human behavior remains largely unexplored. We introduce Be.FM, one of the first open foundation models designed for human behavior modeling. Built upon open-source large language models and fine-tuned on a diverse range of behavioral data, Be.FM can be used to understand and predict human decision-making. We construct a comprehensive set of benchmark tasks for testing the capabilities of behavioral foundation models. Our results demonstrate that Be.FM can predict behaviors, infer characteristics of individuals and populations, generate insights about contexts, and apply behavioral science knowledge.
Authors: Amit Kumthekar, Zion Tilley, Henry Duong, Bhargav Patel, Michael Magnoli, Ahmed Omar, Ahmed Nasser, Chaitanya Gharpure, Yevgen Reztzov
Abstract: Despite the growing clinical adoption of large language models (LLMs), current approaches heavily rely on single model architectures. To overcome risks of obsolescence and rigid dependence on single model systems, we present a novel framework, termed the Consensus Mechanism. Mimicking clinical triage and multidisciplinary clinical decision-making, the Consensus Mechanism implements an ensemble of specialized medical expert agents enabling improved clinical decision making while maintaining robust adaptability. This architecture enables the Consensus Mechanism to be optimized for cost, latency, or performance, purely based on its interior model configuration. To rigorously evaluate the Consensus Mechanism, we employed three medical evaluation benchmarks: MedMCQA, MedQA, and MedXpertQA Text, and the differential diagnosis dataset, DDX+. On MedXpertQA, the Consensus Mechanism achieved an accuracy of 61.0% compared to 53.5% and 45.9% for OpenAI's O3 and Google's Gemini 2.5 Pro. Improvement was consistent across benchmarks with an increase in accuracy on MedQA ($\Delta\mathrm{Accuracy}_{\mathrm{consensus\text{-}O3}} = 3.4\%$) and MedMCQA ($\Delta\mathrm{Accuracy}_{\mathrm{consensus\text{-}O3}} = 9.1\%$). These accuracy gains extended to differential diagnosis generation, where our system demonstrated improved recall and precision (F1$_\mathrm{consensus}$ = 0.326 vs. F1$_{\mathrm{O3\text{-}high}}$ = 0.2886) and a higher top-1 accuracy for DDX (Top1$_\mathrm{consensus}$ = 52.0% vs. Top1$_{\mathrm{O3\text{-}high}}$ = 45.2%).
Authors: Zeyu Liu, Yuhang Liu, Guanghao Zhu, Congkai Xie, Zhen Li, Jianbo Yuan, Xinyao Wang, Qing Li, Shing-Chi Cheung, Shengyu Zhang, Fei Wu, Hongxia Yang
Abstract: Recent advancements in large language models (LLMs) have demonstrated substantial progress in reasoning capabilities, such as DeepSeek-R1, which leverages rule-based reinforcement learning to enhance logical reasoning significantly. However, extending these achievements to multimodal large language models (MLLMs) presents critical challenges, which are frequently more pronounced for Multimodal Small Language Models (MSLMs) given their typically weaker foundational reasoning abilities: (1) the scarcity of high-quality multimodal reasoning datasets, (2) the degradation of reasoning capabilities due to the integration of visual processing, and (3) the risk that direct application of reinforcement learning may produce complex yet incorrect reasoning processes. To address these challenges, we design a novel framework Infi-MMR to systematically unlock the reasoning potential of MSLMs through a curriculum of three carefully structured phases and propose our multimodal reasoning model Infi-MMR-3B. The first phase, Foundational Reasoning Activation, leverages high-quality textual reasoning datasets to activate and strengthen the model's logical reasoning capabilities. The second phase, Cross-Modal Reasoning Adaptation, utilizes caption-augmented multimodal data to facilitate the progressive transfer of reasoning skills to multimodal contexts. The third phase, Multimodal Reasoning Enhancement, employs curated, caption-free multimodal data to mitigate linguistic biases and promote robust cross-modal reasoning. Infi-MMR-3B achieves both state-of-the-art multimodal math reasoning ability (43.68% on MathVerse testmini, 27.04% on MathVision test, and 21.33% on OlympiadBench) and general reasoning ability (67.2% on MathVista testmini).
Authors: Fan Wang, Shaoshan Liu
Abstract: Collective Adaptive Intelligence (CAI) represent a transformative approach in artificial intelligence, wherein numerous autonomous agents collaborate, adapt, and self-organize to navigate complex, dynamic environments. This paradigm is particularly impactful in embodied AI applications, where adaptability and resilience are paramount. By enabling systems to reconfigure themselves in response to unforeseen challenges, CAI facilitate robust performance in real-world scenarios. This article introduces a conceptual framework for designing and analyzing CAI. It delineates key attributes including task generalization, resilience, scalability, and self-assembly, aiming to bridge theoretical foundations with practical methodologies for engineering adaptive, emergent intelligence. By providing a structured foundation for understanding and implementing CAI, this work seeks to guide researchers and practitioners in developing more resilient, scalable, and adaptable AI systems across various domains.
Authors: Mislav Balunovi\'c, Jasper Dekoninck, Ivo Petrov, Nikola Jovanovi\'c, Martin Vechev
Abstract: The rapid advancement of reasoning capabilities in large language models (LLMs) has led to notable improvements on mathematical benchmarks. However, many of the most commonly used evaluation datasets (e.g., AIME 2024) are widely available online, making it difficult to disentangle genuine reasoning from potential memorization. Furthermore, these benchmarks do not evaluate proof-writing capabilities, which are crucial for many mathematical tasks. To address this, we introduce MathArena, a new benchmark based on the following key insight: recurring math competitions provide a stream of high-quality, challenging problems that can be used for real-time evaluation of LLMs. By evaluating models as soon as new problems are released, we effectively eliminate the risk of contamination. Using this framework, we find strong signs of contamination in AIME 2024. Nonetheless, evaluations on harder competitions, such as SMT 2025 -- published well after model release dates -- demonstrate impressive reasoning capabilities in top-performing models. MathArena is also the first benchmark for proof-writing capabilities. On USAMO 2025, even top models score below 25%, far behind their performance on final-answer tasks. So far, we have evaluated 30 models across five competitions, totaling 149 problems. As an evolving benchmark, MathArena will continue to track the progress of LLMs on newly released competitions, ensuring rigorous and up-to-date evaluation of mathematical reasoning.
Authors: Bowen Ping, Minnan Luo, Zhuohang Dang, Chenxi Wang, Chengyou Jia
Abstract: Geometry problem solving presents distinctive challenges in artificial intelligence, requiring exceptional multimodal comprehension and rigorous mathematical reasoning capabilities. Existing approaches typically fall into two categories: neural-based and symbolic-based methods, both of which exhibit limitations in reliability and interpretability. To address this challenge, we propose AutoGPS, a neuro-symbolic collaborative framework that solves geometry problems with concise, reliable, and human-interpretable reasoning processes. Specifically, AutoGPS employs a Multimodal Problem Formalizer (MPF) and a Deductive Symbolic Reasoner (DSR). The MPF utilizes neural cross-modal comprehension to translate geometry problems into structured formal language representations, with feedback from DSR collaboratively. The DSR takes the formalization as input and formulates geometry problem solving as a hypergraph expansion task, executing mathematically rigorous and reliable derivation to produce minimal and human-readable stepwise solutions. Extensive experimental evaluations demonstrate that AutoGPS achieves state-of-the-art performance on benchmark datasets. Furthermore, human stepwise-reasoning evaluation confirms AutoGPS's impressive reliability and interpretability, with 99\% stepwise logical coherence. The project homepage is at https://jayce-ping.github.io/AutoGPS-homepage.
Authors: Ahmad Mohsin, Helge Janicke, Ahmed Ibrahim, Iqbal H. Sarker, Seyit Camtepe
Abstract: This article presents a structured framework for Human-AI collaboration in Security Operations Centers (SOCs), integrating AI autonomy, trust calibration, and Human-in-the-loop decision making. Existing frameworks in SOCs often focus narrowly on automation, lacking systematic structures to manage human oversight, trust calibration, and scalable autonomy with AI. Many assume static or binary autonomy settings, failing to account for the varied complexity, criticality, and risk across SOC tasks considering Humans and AI collaboration. To address these limitations, we propose a novel autonomy tiered framework grounded in five levels of AI autonomy from manual to fully autonomous, mapped to Human-in-the-Loop (HITL) roles and task-specific trust thresholds. This enables adaptive and explainable AI integration across core SOC functions, including monitoring, protection, threat detection, alert triage, and incident response. The proposed framework differentiates itself from previous research by creating formal connections between autonomy, trust, and HITL across various SOC levels, which allows for adaptive task distribution according to operational complexity and associated risks. The framework is exemplified through a simulated cyber range that features the cybersecurity AI-Avatar, a fine-tuned LLM-based SOC assistant. The AI-Avatar case study illustrates human-AI collaboration for SOC tasks, reducing alert fatigue, enhancing response coordination, and strategically calibrating trust. This research systematically presents both the theoretical and practical aspects and feasibility of designing next-generation cognitive SOCs that leverage AI not to replace but to enhance human decision-making.
Authors: Jusheng Zhang, Yijia Fan, Wenjun Lin, Ruiqi Chen, Haoyi Jiang, Wenhao Chai, Jian Wang, Keze Wang
Abstract: We propose GAM-Agent, a game-theoretic multi-agent framework for enhancing vision-language reasoning. Unlike prior single-agent or monolithic models, GAM-Agent formulates the reasoning process as a non-zero-sum game between base agents--each specializing in visual perception subtasks--and a critical agent that verifies logic consistency and factual correctness. Agents communicate via structured claims, evidence, and uncertainty estimates. The framework introduces an uncertainty-aware controller to dynamically adjust agent collaboration, triggering multi-round debates when disagreement or ambiguity is detected. This process yields more robust and interpretable predictions. Experiments on four challenging benchmarks--MMMU, MMBench, MVBench, and V*Bench--demonstrate that GAM-Agent significantly improves performance across various VLM backbones. Notably, GAM-Agent boosts the accuracy of small-to-mid scale models (e.g., Qwen2.5-VL-7B, InternVL3-14B) by 5--6\%, and still enhances strong models like GPT-4o by up to 2--3\%. Our approach is modular, scalable, and generalizable, offering a path toward reliable and explainable multi-agent multimodal reasoning.
Authors: Elisa Celis, Lingxiao Huang, Nisheeth K. Vishnoi
Abstract: The rapid rise of Generative AI (GenAI) tools has sparked debate over their role in complementing or replacing human workers across job contexts. We present a mathematical framework that models jobs, workers, and worker-job fit, introducing a novel decomposition of skills into decision-level and action-level subskills to reflect the complementary strengths of humans and GenAI. We analyze how changes in subskill abilities affect job success, identifying conditions for sharp transitions in success probability. We also establish sufficient conditions under which combining workers with complementary subskills significantly outperforms relying on a single worker. This explains phenomena such as productivity compression, where GenAI assistance yields larger gains for lower-skilled workers. We demonstrate the framework' s practicality using data from O*NET and Big-Bench Lite, aligning real-world data with our model via subskill-division methods. Our results highlight when and how GenAI complements human skills, rather than replacing them.
Authors: Daniel Jarne Ornia, Nicholas Bishop, Joel Dyer, Wei-Chen Lee, Ani Calinescu, Doyne Farme, Michael Wooldridge
Abstract: Advanced reasoning models with agentic capabilities (AI agents) are deployed to interact with humans and to solve sequential decision-making problems under (approximate) utility functions and internal models. When such problems have resource or failure constraints where action sequences may be forcibly terminated once resources are exhausted, agents face implicit trade-offs that reshape their utility-driven (rational) behaviour. Additionally, since these agents are typically commissioned by a human principal to act on their behalf, asymmetries in constraint exposure can give rise to previously unanticipated misalignment between human objectives and agent incentives. We formalise this setting through a survival bandit framework, provide theoretical and empirical results that quantify the impact of survival-driven preference shifts, identify conditions under which misalignment emerges and propose mechanisms to mitigate the emergence of risk-seeking or risk-averse behaviours. As a result, this work aims to increase understanding and interpretability of emergent behaviours of AI agents operating under such survival pressure, and offer guidelines for safely deploying such AI systems in critical resource-limited environments.
Authors: Xiaorui Wu, Xiaofeng Mao, Fei Li, Xin Zhang, Xiaolu Zhang, Jun Zhou, Yuxiang Peng, Li Zheng, Chong Teng, Donghong Ji, Zhuang Li
Abstract: Large language models (LLMs) frequently refuse to respond to pseudo-malicious instructions: semantically harmless input queries triggering unnecessary LLM refusals due to conservative safety alignment, significantly impairing user experience. Collecting such instructions is crucial for evaluating and mitigating over-refusals, but existing instruction curation methods, like manual creation or instruction rewriting, either lack scalability or fail to produce sufficiently diverse and effective refusal-inducing prompts. To address these limitations, we introduce EVOREFUSE, a prompt optimization approach that generates diverse pseudo-malicious instructions consistently eliciting confident refusals across LLMs. EVOREFUSE employs an evolutionary algorithm exploring the instruction space in more diverse directions than existing methods via mutation strategies and recombination, and iteratively evolves seed instructions to maximize evidence lower bound on LLM refusal probability. Using EVOREFUSE, we create two novel datasets: EVOREFUSE-TEST, a benchmark of 582 pseudo-malicious instructions that outperforms the next-best benchmark with 140.41% higher average refusal triggering rate across 9 LLMs, 34.86% greater lexical diversity, and 40.03% improved LLM response confidence scores; and EVOREFUSE-ALIGN, which provides 3,000 pseudo-malicious instructions with responses for supervised and preference-based alignment training. LLAMA3.1-8B-INSTRUCT supervisedly fine-tuned on EVOREFUSE-ALIGN achieves up to 14.31% fewer over-refusals than models trained on the second-best alignment dataset, without compromising safety. Our analysis with EVOREFUSE-TEST reveals models trigger over-refusals by overly focusing on sensitive keywords while ignoring broader context.
Authors: Xiang Li, Haiyang Yu, Xinghua Zhang, Ziyang Huang, Shizhu He, Kang Liu, Jun Zhao, Fei Huang, Yongbin Li
Abstract: Process Reward Models (PRMs) are crucial in complex reasoning and problem-solving tasks (e.g., LLM agents with long-horizon decision-making) by verifying the correctness of each intermediate reasoning step. In real-world scenarios, LLMs may apply various reasoning patterns (e.g., decomposition) to solve a problem, potentially suffering from errors under various reasoning patterns. Therefore, PRMs are required to identify errors under various reasoning patterns during the reasoning process. However, existing benchmarks mainly focus on evaluating PRMs with stepwise correctness, ignoring a systematic evaluation of PRMs under various reasoning patterns. To mitigate this gap, we introduce Socratic-PRMBench, a new benchmark to evaluate PRMs systematically under six reasoning patterns, including Transformation, Decomposition, Regather, Deduction, Verification, and Integration. Socratic-PRMBench}comprises 2995 reasoning paths with flaws within the aforementioned six reasoning patterns. Through our experiments on both PRMs and LLMs prompted as critic models, we identify notable deficiencies in existing PRMs. These observations underscore the significant weakness of current PRMs in conducting evaluations on reasoning steps under various reasoning patterns. We hope Socratic-PRMBench can serve as a comprehensive testbed for systematic evaluation of PRMs under diverse reasoning patterns and pave the way for future development of PRMs.
Authors: Ke Weng, Lun Du, Sirui Li, Wangyue Lu, Haozhe Sun, Hengyu Liu, Tiancheng Zhang
Abstract: Autoformalization, the process of transforming informal mathematical propositions into verifiable formal representations, is a foundational task in automated theorem proving, offering a new perspective on the use of mathematics in both theoretical and applied domains. Driven by the rapid progress in artificial intelligence, particularly large language models (LLMs), this field has witnessed substantial growth, bringing both new opportunities and unique challenges. In this survey, we provide a comprehensive overview of recent advances in autoformalization from both mathematical and LLM-centric perspectives. We examine how autoformalization is applied across various mathematical domains and levels of difficulty, and analyze the end-to-end workflow from data preprocessing to model design and evaluation. We further explore the emerging role of autoformalization in enhancing the verifiability of LLM-generated outputs, highlighting its potential to improve both the trustworthiness and reasoning capabilities of LLMs. Finally, we summarize key open-source models and datasets supporting current research, and discuss open challenges and promising future directions for the field.
Authors: Hangoo Kang, Jehyeok Yeon, Gagandeep Singh
Abstract: Autonomous agentic AI systems powered by vision-language models (VLMs) are rapidly advancing toward real-world deployment, yet their cross-modal reasoning capabilities introduce new attack surfaces for adversarial manipulation that exploit semantic reasoning across modalities. Existing adversarial attacks typically rely on visible pixel perturbations or require privileged model or environment access, making them impractical for stealthy, real-world exploitation. We introduce TRAP, a generative adversarial framework that manipulates the agent's decision-making using diffusion-based semantic injections. Our method combines negative prompt-based degradation with positive semantic optimization, guided by a Siamese semantic network and layout-aware spatial masking. Without requiring access to model internals, TRAP produces visually natural images yet induces consistent selection biases in agentic AI systems. We evaluate TRAP on the Microsoft Common Objects in Context (COCO) dataset, building multi-candidate decision scenarios. Across these scenarios, TRAP achieves a 100% attack success rate on leading models, including LLaVA-34B, Gemma3, and Mistral-3.1, significantly outperforming baselines such as SPSA, Bandit, and standard diffusion approaches. These results expose a critical vulnerability: Autonomous agents can be consistently misled through human-imperceptible cross-modal manipulations. These findings highlight the need for defense strategies beyond pixel-level robustness to address semantic vulnerabilities in cross-modal decision-making.
Authors: Ruiqi He, Falk Lieder
Abstract: People employ efficient planning strategies. But how are these strategies acquired? Previous research suggests that people can discover new planning strategies through learning from reinforcements, a process known as metacognitive reinforcement learning (MCRL). While prior work has shown that MCRL models can learn new planning strategies and explain more participants' experience-driven discovery better than alternative mechanisms, it also revealed significant individual differences in metacognitive learning. Furthermore, when fitted to human data, these models exhibit a slower rate of strategy discovery than humans. In this study, we investigate whether incorporating cognitive mechanisms that might facilitate human strategy discovery can bring models of MCRL closer to human performance. Specifically, we consider intrinsically generated metacognitive pseudo-rewards, subjective effort valuation, and termination deliberation. Analysis of planning task data shows that a larger proportion of participants used at least one of these mechanisms, with significant individual differences in their usage and varying impacts on strategy discovery. Metacognitive pseudo-rewards, subjective effort valuation, and learning the value of acting without further planning were found to facilitate strategy discovery. While these enhancements provided valuable insights into individual differences and the effect of these mechanisms on strategy discovery, they did not fully close the gap between model and human performance, prompting further exploration of additional factors that people might use to discover new planning strategies.
Authors: Janik-Vasily Benzin, Gyunam Park, Stefanie Rinderle-Ma
Abstract: Model abstraction (MA) and event abstraction (EA) are means to reduce complexity of (discovered) models and event data. Imagine a process intelligence project that aims to analyze a model discovered from event data which is further abstracted, possibly multiple times, to reach optimality goals, e.g., reducing model size. So far, after discovering the model, there is no technique that enables the synchronized abstraction of the underlying event log. This results in loosing the grounding in the real-world behavior contained in the log and, in turn, restricts analysis insights. Hence, in this work, we provide the formal basis for synchronized model and event abstraction, i.e., we prove that abstracting a process model by MA and discovering a process model from an abstracted event log yields an equivalent process model. We prove the feasibility of our approach based on behavioral profile abstraction as non-order preserving MA technique, resulting in a novel EA technique.
Authors: Kunlun Zhu, Jiaxun Zhang, Ziheng Qi, Nuoxing Shang, Zijia Liu, Peixuan Han, Yue Su, Haofei Yu, Jiaxuan You
Abstract: Recent advancements in large language model (LLM) agents have significantly accelerated scientific discovery automation, yet concurrently raised critical ethical and safety concerns. To systematically address these challenges, we introduce \textbf{SafeScientist}, an innovative AI scientist framework explicitly designed to enhance safety and ethical responsibility in AI-driven scientific exploration. SafeScientist proactively refuses ethically inappropriate or high-risk tasks and rigorously emphasizes safety throughout the research process. To achieve comprehensive safety oversight, we integrate multiple defensive mechanisms, including prompt monitoring, agent-collaboration monitoring, tool-use monitoring, and an ethical reviewer component. Complementing SafeScientist, we propose \textbf{SciSafetyBench}, a novel benchmark specifically designed to evaluate AI safety in scientific contexts, comprising 240 high-risk scientific tasks across 6 domains, alongside 30 specially designed scientific tools and 120 tool-related risk tasks. Extensive experiments demonstrate that SafeScientist significantly improves safety performance by 35\% compared to traditional AI scientist frameworks, without compromising scientific output quality. Additionally, we rigorously validate the robustness of our safety pipeline against diverse adversarial attack methods, further confirming the effectiveness of our integrated approach. The code and data will be available at https://github.com/ulab-uiuc/SafeScientist. \textcolor{red}{Warning: this paper contains example data that may be offensive or harmful.}
Authors: Benjamin Arnav, Pablo Bernabeu-P\'erez, Nathan Helm-Burger, Tim Kostolansky, Hannes Whittingham, Mary Phuong
Abstract: As AI models are deployed with increasing autonomy, it is important to ensure they do not take harmful actions unnoticed. As a potential mitigation, we investigate Chain-of-Thought (CoT) monitoring, wherein a weaker trusted monitor model continuously oversees the intermediate reasoning steps of a more powerful but untrusted model. We compare CoT monitoring to action-only monitoring, where only final outputs are reviewed, in a red-teaming setup where the untrusted model is instructed to pursue harmful side tasks while completing a coding problem. We find that CoT monitoring improves detection by up to 27 percentage points in scenarios where action-only monitoring fails to reliably identify sabotage. However, CoT traces can also contain misleading rationalizations that deceive the monitor, reducing performance in more obvious sabotage cases. To address this, we introduce a hybrid protocol that independently scores both reasoning and final outputs and combines them using a weighted average. This hybrid monitor consistently outperforms both CoT and action-only monitors across all tested models and tasks, with detection rates over four times higher than action-only monitoring for subtle deception scenarios.
Authors: Linqiang Guo (Peter), Wei Liu (Peter), Yi Wen Heng (Peter), Tse-Hsun (Peter), Chen, Yang Wang
Abstract: Mobile GUI agents aim to autonomously complete user-instructed tasks across mobile apps. Recent advances in Multimodal Large Language Models (MLLMs) enable these agents to interpret UI screens, identify actionable elements, and perform interactions such as tapping or typing. However, existing agents remain reactive: they reason only over the current screen and lack a structured model of app navigation flow, limiting their ability to understand context, detect unexpected outcomes, and recover from errors. We present MAPLE, a state-aware multi-agent framework that abstracts app interactions as a Finite State Machine (FSM). We computationally model each UI screen as a discrete state and user actions as transitions, allowing the FSM to provide a structured representation of the app execution. MAPLE consists of specialized agents responsible for four phases of task execution: planning, execution, verification, error recovery, and knowledge retention. These agents collaborate to dynamically construct FSMs in real time based on perception data extracted from the UI screen, allowing the GUI agents to track navigation progress and flow, validate action outcomes through pre- and post-conditions of the states, and recover from errors by rolling back to previously stable states. Our evaluation results on two challenging cross-app benchmarks, Mobile-Eval-E and SPA-Bench, show that MAPLE outperforms the state-of-the-art baseline, improving task success rate by up to 12%, recovery success by 13.8%, and action accuracy by 6.5%. Our results highlight the importance of structured state modeling in guiding mobile GUI agents during task execution. Moreover, our FSM representation can be integrated into future GUI agent architectures as a lightweight, model-agnostic memory layer to support structured planning, execution verification, and error recovery.
Authors: Lang Cao, Jingxian Xu, Hanbing Liu, Jinyu Wang, Mengyu Zhou, Haoyu Dong, Shi Han, Dongmei Zhang
Abstract: Tables are a fundamental structure for organizing and analyzing data, making effective table understanding a critical capability for intelligent systems. While large language models (LMs) demonstrate strong general reasoning abilities, they continue to struggle with accurate numerical or symbolic reasoning over tabular data, especially in complex scenarios. Spreadsheet formulas provide a powerful and expressive medium for representing executable symbolic operations, encoding rich reasoning patterns that remain largely underutilized. In this paper, we propose Formula Tuning (Fortune), a reinforcement learning (RL) framework that trains LMs to generate executable spreadsheet formulas for question answering over general tabular data. Formula Tuning reduces the reliance on supervised formula annotations by using binary answer correctness as a reward signal, guiding the model to learn formula derivation through reasoning. We provide a theoretical analysis of its advantages and demonstrate its effectiveness through extensive experiments on seven table reasoning benchmarks. Formula Tuning substantially enhances LM performance, particularly on multi-step numerical and symbolic reasoning tasks, enabling a 7B model to outperform O1 on table understanding. This highlights the potential of formula-driven RL to advance symbolic table reasoning in LMs.
Authors: Caroline Wang, Arrasy Rahman, Jiaxun Cui, Yoonchang Sung, Peter Stone
Abstract: Developing AI agents capable of collaborating with previously unseen partners is a fundamental generalization challenge in multi-agent learning, known as Ad Hoc Teamwork (AHT). Existing AHT approaches typically adopt a two-stage pipeline, where first, a fixed population of teammates is generated with the idea that they should be representative of the teammates that will be seen at deployment time, and second, an AHT agent is trained to collaborate well with agents in the population. To date, the research community has focused on designing separate algorithms for each stage. This separation has led to algorithms that generate teammate pools with limited coverage of possible behaviors, and that ignore whether the generated teammates are easy to learn from for the AHT agent. Furthermore, algorithms for training AHT agents typically treat the set of training teammates as static, thus attempting to generalize to previously unseen partner agents without assuming any control over the distribution of training teammates. In this paper, we present a unified framework for AHT by reformulating the problem as an open-ended learning process between an ad hoc agent and an adversarial teammate generator. We introduce ROTATE, a regret-driven, open-ended training algorithm that alternates between improving the AHT agent and generating teammates that probe its deficiencies. Extensive experiments across diverse AHT environments demonstrate that ROTATE significantly outperforms baselines at generalizing to an unseen set of evaluation teammates, thus establishing a new standard for robust and generalizable teamwork.
Authors: Ran Zhang, Mohannad Elhamod
Abstract: The rapid advancement of LLMs has led to the creation of diverse agentic systems in data analysis, utilizing LLMs' capabilities to improve insight generation and visualization. In this paper, we present an agentic system that automates the data-to-dashboard pipeline through modular LLM agents capable of domain detection, concept extraction, multi-perspective analysis generation, and iterative self-reflection. Unlike existing chart QA systems, our framework simulates the analytical reasoning process of business analysts by retrieving domain-relevant knowledge and adapting to diverse datasets without relying on closed ontologies or question templates. We evaluate our system on three datasets across different domains. Benchmarked against GPT-4o with a single-prompt baseline, our approach shows improved insightfulness, domain relevance, and analytical depth, as measured by tailored evaluation metrics and qualitative human assessment. This work contributes a novel modular pipeline to bridge the path from raw data to visualization, and opens new opportunities for human-in-the-loop validation by domain experts in business analytics. All code can be found here: https://github.com/77luvC/D2D_Data2Dashboard
Authors: Ruida Wang (May), Yuxin Li (May), Yi R. (May), Fung, Tong Zhang
Abstract: Enhancing the mathematical reasoning capabilities of LLMs has garnered significant attention in both the mathematical and computer science communities. Recent works have made substantial progress in both Natural Language (NL) reasoning and Formal Language (FL) reasoning by leveraging the potential of pure Reinforcement Learning (RL) methods on base models. However, RL approaches struggle to impart new capabilities not presented in the base model, highlighting the need to integrate more knowledge like FL into NL math reasoning effectively. Yet, this integration is challenging due to inherent disparities in problem structure and reasoning format between NL and FL. To address these challenges, we introduce **NL-FL HybridReasoning**, an end-to-end framework designed to incorporate the FL expert into NL math problem-solving. To bridge the NL and FL input format gap, we propose the *NL-FL Problem Alignment* method, which reformulates the Question-Answering (QA) problems in NL as existence theorems in FL. Subsequently, the *Mixed Problem Input* technique we provide enables the FL reasoner to handle both QA and existence problems concurrently. Lastly, we mitigate the NL and FL output format gap in reasoning through an LLM-based *Answer Extraction* mechanism. Comprehensive experiments demonstrate that the **HybridReasoning** framework achieves **89.80%** and **84.34%** accuracy rates on the MATH-500 and the AMC benchmarks, surpassing the NL baseline by 4.60% and 4.82%, respectively. Notably, some problems resolved by our framework remain unsolved by the NL baseline model even under a larger number of trials.
Authors: Hugo Henry, Kelly Cohen
Abstract: This study investigates the application of Genetic Fuzzy Systems (GFS) to model the self-noise generated by airfoils, a key issue in aeroaccoustics with significant implications for aerospace, automotive and drone applications. Using the publicly available Airfoil Self Noise dataset, various Fuzzy regression strategies are explored and compared. The paper evaluates a brute force Takagi Sugeno Kang (TSK) fuzzy system with high rule density, a cascading Geneti Fuzzy Tree (GFT) architecture and a novel clustered approach based on Fuzzy C-means (FCM) to reduce the model's complexity. This highlights the viability of clustering assisted fuzzy inference as an effective regression tool for complex aero accoustic phenomena. Keywords : Fuzzy logic, Regression, Cascading systems, Clustering and AI.
Authors: Chenyu Yang, Shiqian Su, Shi Liu, Xuan Dong, Yue Yu, Weijie Su, Xuehui Wang, Zhaoyang Liu, Jinguo Zhu, Hao Li, Wenhai Wang, Yu Qiao, Xizhou Zhu, Jifeng Dai
Abstract: The rapid advancement of large Vision-Language Models (VLMs) has propelled the development of pure-vision-based GUI Agents, capable of perceiving and operating Graphical User Interfaces (GUI) to autonomously fulfill user instructions. However, existing approaches usually adopt an offline learning framework, which faces two core limitations: (1) heavy reliance on high-quality manual annotations for element grounding and action supervision, and (2) limited adaptability to dynamic and interactive environments. To address these limitations, we propose ZeroGUI, a scalable, online learning framework for automating GUI Agent training at Zero human cost. Specifically, ZeroGUI integrates (i) VLM-based automatic task generation to produce diverse training goals from the current environment state, (ii) VLM-based automatic reward estimation to assess task success without hand-crafted evaluation functions, and (iii) two-stage online reinforcement learning to continuously interact with and learn from GUI environments. Experiments on two advanced GUI Agents (UI-TARS and Aguvis) demonstrate that ZeroGUI significantly boosts performance across OSWorld and AndroidLab environments. The code is available at https://github.com/OpenGVLab/ZeroGUI.
Authors: Wasif Khan, Kyle B. See, Simon Kato, Ziqian Huang, Amy Lazarte, Kyle Douglas, Xiangyang Lou, Teng J. Peng, Dhanashree Rajderkar, John Rees, Pina Sanelli, Amita Singh, Ibrahim Tuna, Christina A. Wilson, Ruogu Fang
Abstract: Perfusion imaging is extensively utilized to assess hemodynamic status and tissue perfusion in various organs. Computed tomography perfusion (CTP) imaging plays a key role in the early assessment and planning of stroke treatment. While CTP provides essential perfusion parameters to identify abnormal blood flow in the brain, the use of contrast agents in CTP can lead to allergic reactions and adverse side effects, along with costing USD 4.9 billion worldwide in 2022. To address these challenges, we propose a novel deep learning framework called Multitask Automated Generation of Intermodal CT perfusion maps (MAGIC). This framework combines generative artificial intelligence and physiological information to map non-contrast computed tomography (CT) imaging to multiple contrast-free CTP imaging maps. We demonstrate enhanced image fidelity by incorporating physiological characteristics into the loss terms. Our network was trained and validated using CT image data from patients referred for stroke at UF Health and demonstrated robustness to abnormalities in brain perfusion activity. A double-blinded study was conducted involving seven experienced neuroradiologists and vascular neurologists. This study validated MAGIC's visual quality and diagnostic accuracy showing favorable performance compared to clinical perfusion imaging with intravenous contrast injection. Overall, MAGIC holds great promise in revolutionizing healthcare by offering contrast-free, cost-effective, and rapid perfusion imaging.
Authors: Pawan Neupane, Jian Liu, Jianlin Cheng
Abstract: Predicting protein complex structures is essential for protein function analysis, protein design, and drug discovery. While AI methods like AlphaFold can predict accurate structural models for many protein complexes, reliably estimating the quality of these predicted models (estimation of model accuracy, or EMA) for model ranking and selection remains a major challenge. A key barrier to developing effective machine learning-based EMA methods is the lack of large, diverse, and well-annotated datasets for training and evaluation. To address this gap, we introduce PSBench, a benchmark suite comprising four large-scale, labeled datasets generated during the 15th and 16th community-wide Critical Assessment of Protein Structure Prediction (CASP15 and CASP16). PSBench includes over one million structural models covering a wide range of protein sequence lengths, complex stoichiometries, functional classes, and modeling difficulties. Each model is annotated with multiple complementary quality scores at the global, local, and interface levels. PSBench also provides multiple evaluation metrics and baseline EMA methods to facilitate rigorous comparisons. To demonstrate PSBench's utility, we trained and evaluated GATE, a graph transformer-based EMA method, on the CASP15 data. GATE was blindly tested in CASP16 (2024), where it ranked among the top-performing EMA methods. These results highlight PSBench as a valuable resource for advancing EMA research in protein complex modeling. PSBench is publicly available at: https://github.com/BioinfoMachineLearning/PSBench.
Authors: Xuhang Chen, Michael Kwok-Po Ng, Kim-Fung Tsang, Chi-Man Pun, Shuqiang Wang
Abstract: Brain network analysis plays a crucial role in diagnosing and monitoring neurodegenerative disorders such as Alzheimer's disease (AD). Existing approaches for constructing structural brain networks from diffusion tensor imaging (DTI) often rely on specialized toolkits that suffer from inherent limitations: operator subjectivity, labor-intensive workflows, and restricted capacity to capture complex topological features and disease-specific biomarkers. To overcome these challenges and advance computational neuroimaging instrumentation, ConnectomeDiffuser is proposed as a novel diffusion-based framework for automated end-to-end brain network construction from DTI. The proposed model combines three key components: (1) a Template Network that extracts topological features from 3D DTI scans using Riemannian geometric principles, (2) a diffusion model that generates comprehensive brain networks with enhanced topological fidelity, and (3) a Graph Convolutional Network classifier that incorporates disease-specific markers to improve diagnostic accuracy. ConnectomeDiffuser demonstrates superior performance by capturing a broader range of structural connectivity and pathology-related information, enabling more sensitive analysis of individual variations in brain networks. Experimental validation on datasets representing two distinct neurodegenerative conditions demonstrates significant performance improvements over other brain network methods. This work contributes to the advancement of instrumentation in the context of neurological disorders, providing clinicians and researchers with a robust, generalizable measurement framework that facilitates more accurate diagnosis, deeper mechanistic understanding, and improved therapeutic monitoring of neurodegenerative diseases such as AD.
Authors: Marcus J. Vroemen, Yuqian Chen, Yui Lo, Tengfei Xu, Weidong Cai, Fan Zhang, Josien P. W. Pluim, Lauren J. O'Donnell
Abstract: Diffusion MRI (dMRI) tractography enables in vivo mapping of brain structural connections, but traditional connectome generation is time-consuming and requires gray matter parcellation, posing challenges for large-scale studies. We introduce DeepMultiConnectome, a deep-learning model that predicts structural connectomes directly from tractography, bypassing the need for gray matter parcellation while supporting multiple parcellation schemes. Using a point-cloud-based neural network with multi-task learning, the model classifies streamlines according to their connected regions across two parcellation schemes, sharing a learned representation. We train and validate DeepMultiConnectome on tractography from the Human Connectome Project Young Adult dataset ($n = 1000$), labeled with an 84 and 164 region gray matter parcellation scheme. DeepMultiConnectome predicts multiple structural connectomes from a whole-brain tractogram containing 3 million streamlines in approximately 40 seconds. DeepMultiConnectome is evaluated by comparing predicted connectomes with traditional connectomes generated using the conventional method of labeling streamlines using a gray matter parcellation. The predicted connectomes are highly correlated with traditionally generated connectomes ($r = 0.992$ for an 84-region scheme; $r = 0.986$ for a 164-region scheme) and largely preserve network properties. A test-retest analysis of DeepMultiConnectome demonstrates reproducibility comparable to traditionally generated connectomes. The predicted connectomes perform similarly to traditionally generated connectomes in predicting age and cognitive function. Overall, DeepMultiConnectome provides a scalable, fast model for generating subject-specific connectomes across multiple parcellation schemes.
Authors: Feng Yao, Zilong Wang, Liyuan Liu, Junxia Cui, Li Zhong, Xiaohan Fu, Haohui Mai, Vish Krishnan, Jianfeng Gao, Jingbo Shang
Abstract: Code generation with large language models (LLMs), often termed vibe coding, is increasingly adopted in production but fails to ensure code quality, particularly in security (e.g., SQL injection vulnerabilities) and maintainability (e.g., missing type annotations). Existing methods, such as supervised fine-tuning and rule-based post-processing, rely on labor-intensive annotations or brittle heuristics, limiting their scalability and effectiveness. We propose REAL, a reinforcement learning framework that incentivizes LLMs to generate production-quality code using program analysis-guided feedback. Specifically, REAL integrates two automated signals: (1) program analysis detecting security or maintainability defects and (2) unit tests ensuring functional correctness. Unlike prior work, our framework is prompt-agnostic and reference-free, enabling scalable supervision without manual intervention. Experiments across multiple datasets and model scales demonstrate that REAL outperforms state-of-the-art methods in simultaneous assessments of functionality and code quality. Our work bridges the gap between rapid prototyping and production-ready code, enabling LLMs to deliver both speed and quality.
Authors: Tamas Spisak, Karl Friston
Abstract: Attractor dynamics are a hallmark of many complex systems, including the brain. Understanding how such self-organizing dynamics emerge from first principles is crucial for advancing our understanding of neuronal computations and the design of artificial intelligence systems. Here we formalize how attractor networks emerge from the free energy principle applied to a universal partitioning of random dynamical systems. Our approach obviates the need for explicitly imposed learning and inference rules and identifies emergent, but efficient and biologically plausible inference and learning dynamics for such self-organizing systems. These result in a collective, multi-level Bayesian active inference process. Attractors on the free energy landscape encode prior beliefs; inference integrates sensory data into posterior beliefs; and learning fine-tunes couplings to minimize long-term surprise. Analytically and via simulations, we establish that the proposed networks favor approximately orthogonalized attractor representations, a consequence of simultaneously optimizing predictive accuracy and model complexity. These attractors efficiently span the input subspace, enhancing generalization and the mutual information between hidden causes and observable effects. Furthermore, while random data presentation leads to symmetric and sparse couplings, sequential data fosters asymmetric couplings and non-equilibrium steady-state dynamics, offering a natural extension to conventional Boltzmann Machines. Our findings offer a unifying theory of self-organizing attractor networks, providing novel insights for AI and neuroscience.
Authors: Ansar Aynetdinov, Alan Akbik
Abstract: Multi-token prediction (MTP) is a recently proposed pre-training objective for language models. Rather than predicting only the next token (NTP), MTP predicts the next $k$ tokens at each prediction step, using multiple prediction heads. MTP has shown promise in improving downstream performance, inference speed, and training efficiency, particularly for large models. However, prior work has shown that smaller language models (SLMs) struggle with the MTP objective. To address this, we propose a curriculum learning strategy for MTP training, exploring two variants: a forward curriculum, which gradually increases the complexity of the pre-training objective from NTP to MTP, and a reverse curriculum, which does the opposite. Our experiments show that the forward curriculum enables SLMs to better leverage the MTP objective during pre-training, improving downstream NTP performance and generative output quality, while retaining the benefits of self-speculative decoding. The reverse curriculum achieves stronger NTP performance and output quality, but fails to provide any self-speculative decoding benefits.
Authors: Sara Papi, Marco Gaido, Luisa Bentivogli, Alessio Brutti, Mauro Cettolo, Roberto Gretter, Marco Matassoni, Mohamed Nabih, Matteo Negri
Abstract: The development of speech foundation models (SFMs) like Whisper and SeamlessM4T has significantly advanced the field of speech processing. However, their closed nature--with inaccessible training data and code--poses major reproducibility and fair evaluation challenges. While other domains have made substantial progress toward open science by developing fully transparent models trained on open-source (OS) code and data, similar efforts in speech remain limited. To fill this gap, we introduce FAMA, the first family of open science SFMs for English and Italian, trained on 150k+ hours of OS speech data. Moreover, we present a new dataset containing 16k hours of cleaned and pseudo-labeled speech for both languages. Results show that FAMA achieves competitive performance compared to existing SFMs while being up to 8 times faster. All artifacts, including code, datasets, and models, are released under OS-compliant licenses, promoting openness in speech technology research.
Authors: Afila Ajithkumar Sophiya, Akarsh K Nair, Sepehr Maleki, Senthil K. Krishnababu
Abstract: Physics Informed Neural Networks (PINNs) have been emerging as a powerful computational tool for solving differential equations. However, the applicability of these models is still in its initial stages and requires more standardization to gain wider popularity. Through this survey, we present a comprehensive overview of PINNs approaches exploring various aspects related to their architecture, variants, areas of application, real-world use cases, challenges, and so on. Even though existing surveys can be identified, they fail to provide a comprehensive view as they primarily focus on either different application scenarios or limit their study to a superficial level. This survey attempts to bridge the gap in the existing literature by presenting a detailed analysis of all these factors combined with recent advancements and state-of-the-art research in PINNs. Additionally, we discuss prevalent challenges in PINNs implementation and present some of the future research directions as well. The overall contributions of the survey can be summarised into three sections: A detailed overview of PINNs architecture and variants, a performance analysis of PINNs on different equations and application domains highlighting their features. Finally, we present a detailed discussion of current issues and future research directions.
Authors: Marco Colussi, Dragan Ahmetovic, Sergio Mascetti
Abstract: This paper presents MIAS-SAM, a novel approach for the segmentation of anomalous regions in medical images. MIAS-SAM uses a patch-based memory bank to store relevant image features, which are extracted from normal data using the SAM encoder. At inference time, the embedding patches extracted from the SAM encoder are compared with those in the memory bank to obtain the anomaly map. Finally, MIAS-SAM computes the center of gravity of the anomaly map to prompt the SAM decoder, obtaining an accurate segmentation from the previously extracted features. Differently from prior works, MIAS-SAM does not require to define a threshold value to obtain the segmentation from the anomaly map. Experimental results conducted on three publicly available datasets, each with a different imaging modality (Brain MRI, Liver CT, and Retina OCT) show accurate anomaly segmentation capabilities measured using DICE score. The code is available at: https://github.com/warpcut/MIAS-SAM
Authors: Eleni Vasilaki
Abstract: Large Language Models (LLMs) are typically analysed through architectural, behavioural, or training-data lenses. This article offers a theoretical and experiential re-framing: LLMs as dynamic instantiations of Collective human Knowledge (CK), where intelligence is evoked through dialogue rather than stored statically. Drawing on concepts from neuroscience and AI, and grounded in sustained interaction with ChatGPT-4, I examine emergent dialogue patterns, the implications of fine-tuning, and the notion of co-augmentation: mutual enhancement between human and machine cognition. This perspective offers a new lens for understanding interaction, representation, and agency in contemporary AI systems.
Authors: Christopher Ormerod
Abstract: This study illustrates how incorporating feedback-oriented annotations into the scoring pipeline can enhance the accuracy of automated essay scoring (AES). This approach is demonstrated with the Persuasive Essays for Rating, Selecting, and Understanding Argumentative and Discourse Elements (PERSUADE) corpus. We integrate two types of feedback-driven annotations: those that identify spelling and grammatical errors, and those that highlight argumentative components. To illustrate how this method could be applied in real-world scenarios, we employ two LLMs to generate annotations -- a generative language model used for spell-correction and an encoder-based token classifier trained to identify and mark argumentative elements. By incorporating annotations into the scoring process, we demonstrate improvements in performance using encoder-based large language models fine-tuned as classifiers.
Authors: Anton Bj\"orklund, Mykola Zaitsev, Marta Kwiatkowska
Abstract: The growing reliance on artificial intelligence in safety- and security-critical applications demands effective neural network certification. A challenging real-world use case is certification against ``patch attacks'', where adversarial patches or lighting conditions obscure parts of images, for example traffic signs. One approach to certification, which also gives quantitative coverage estimates, utilizes preimages of neural networks, i.e., the set of inputs that lead to a specified output. However, these preimage approximation methods, including the state-of-the-art PREMAP algorithm, struggle with scalability. This paper presents novel algorithmic improvements to PREMAP involving tighter bounds, adaptive Monte Carlo sampling, and improved branching heuristics. We demonstrate efficiency improvements of at least an order of magnitude on reinforcement learning control benchmarks, and show that our method scales to convolutional neural networks that were previously infeasible. Our results demonstrate the potential of preimage approximation methodology for reliability and robustness certification.
Authors: Pedro Mendes, Paolo Romano, David Garlan
Abstract: Reliable uncertainty estimation is critical for deploying neural networks (NNs) in real-world applications. While existing calibration techniques often rely on post-hoc adjustments or coarse-grained binning methods, they remain limited in scalability, differentiability, and generalization across domains. In this work, we introduce CLUE (Calibration via Learning Uncertainty-Error Alignment), a novel approach that explicitly aligns predicted uncertainty with observed error during training, grounded in the principle that well-calibrated models should produce uncertainty estimates that match their empirical loss. CLUE adopts a novel loss function that jointly optimizes predictive performance and calibration, using summary statistics of uncertainty and loss as proxies. The proposed method is fully differentiable, domain-agnostic, and compatible with standard training pipelines. Through extensive experiments on vision, regression, and language modeling tasks, including out-of-distribution and domain-shift scenarios, we demonstrate that CLUE achieves superior calibration quality and competitive predictive performance with respect to state-of-the-art approaches without imposing significant computational overhead.
Authors: Andrew Zhu, Evan Osgood, Chris Callison-Burch
Abstract: Much work has been done on conversational LLM agents which directly assist human users with tasks. We present an alternative paradigm for interacting with LLM agents, which we call "overhearing agents". These overhearing agents do not actively participate in conversation -- instead, they "listen in" on human-to-human conversations and perform background tasks or provide suggestions to assist the user. In this work, we explore the overhearing agents paradigm through the lens of Dungeons & Dragons gameplay. We present an in-depth study using large multimodal audio-language models as overhearing agents to assist a Dungeon Master. We perform a human evaluation to examine the helpfulness of such agents and find that some large audio-language models have the emergent ability to perform overhearing agent tasks using implicit audio cues. Finally, we release Python libraries and our project code to support further research into the overhearing agents paradigm at https://github.com/zhudotexe/overhearing_agents.
Authors: Jonghan Lim, Ilya Kovalenko
Abstract: Manufacturing environments are becoming more complex and unpredictable due to factors such as demand variations and shorter product lifespans. This complexity requires real-time decision-making and adaptation to disruptions. Traditional control approaches highlight the need for advanced control strategies capable of overcoming unforeseen challenges, as they demonstrate limitations in responsiveness within dynamic industrial settings. Multi-agent systems address these challenges through decentralization of decision-making, enabling systems to respond dynamically to operational changes. However, current multi-agent systems encounter challenges related to real-time adaptation, context-aware decision-making, and the dynamic exploration of resource capabilities. Large language models provide the possibility to overcome these limitations through context-aware decision-making capabilities. This paper introduces a large language model-enabled control architecture for multi-agent manufacturing systems to dynamically explore resource capabilities in response to real-time disruptions. A simulation-based case study demonstrates that the proposed architecture improves system resilience and flexibility. The case study findings show improved throughput and efficient resource utilization compared to existing approaches.
Authors: Zhangyi Hu, Jiemin Wu, Hua Xu, Mingqian Liao, Ninghui Feng, Bo Gao, Songning Lai, Yutao Yue
Abstract: Irregular Multivariate Time Series (IMTS) forecasting is challenging due to the unaligned nature of multi-channel signals and the prevalence of extensive missing data. Existing methods struggle to capture reliable temporal patterns from such data due to significant missing values. While pre-trained foundation models show potential for addressing these challenges, they are typically designed for Regularly Sampled Time Series (RTS). Motivated by the visual Mask AutoEncoder's (MAE) powerful capability for modeling sparse multi-channel information and its success in RTS forecasting, we propose VIMTS, a framework adapting Visual MAE for IMTS forecasting. To mitigate the effect of missing values, VIMTS first processes IMTS along the timeline into feature patches at equal intervals. These patches are then complemented using learned cross-channel dependencies. Then it leverages visual MAE's capability in handling sparse multichannel data for patch reconstruction, followed by a coarse-to-fine technique to generate precise predictions from focused contexts. In addition, we integrate self-supervised learning for improved IMTS modeling by adapting the visual MAE to IMTS data. Extensive experiments demonstrate VIMTS's superior performance and few-shot capability, advancing the application of visual foundation models in more general time series tasks. Our code is available at https://github.com/WHU-HZY/VIMTS.
Authors: Linghan Zhong, Samuel Yuan, Jiyang Zhang, Yu Liu, Pengyu Nie, Junyi Jessy Li, Milos Gligoric
Abstract: Exceptional behavior tests (EBTs) are crucial in software development for verifying that code correctly handles unwanted events and throws appropriate exceptions. However, prior research has shown that developers often prioritize testing "happy paths", e.g., paths without unwanted events over exceptional scenarios. We present exLong, a framework that automatically generates EBTs to address this gap. exLong leverages a large language model (LLM) fine-tuned from CodeLlama and incorporates reasoning about exception-throwing traces, conditional expressions that guard throw statements, and non-exceptional behavior tests that execute similar traces. Our demonstration video illustrates how exLong can effectively assist developers in creating comprehensive EBTs for their project (available at https://youtu.be/Jro8kMgplZk).
Authors: Michael Klamkin, Mathieu Tanneau, Pascal Van Hentenryck
Abstract: Machine Learning (ML) techniques for Optimal Power Flow (OPF) problems have recently garnered significant attention, reflecting a broader trend of leveraging ML to approximate and/or accelerate the resolution of complex optimization problems. These developments are necessitated by the increased volatility and scale in energy production for modern and future grids. However, progress in ML for OPF is hindered by the lack of standardized datasets and evaluation metrics, from generating and solving OPF instances, to training and benchmarking machine learning models. To address this challenge, this paper introduces PGLearn, a comprehensive suite of standardized datasets and evaluation tools for ML and OPF. PGLearn provides datasets that are representative of real-life operating conditions, by explicitly capturing both global and local variability in the data generation, and by, for the first time, including time series data for several large-scale systems. In addition, it supports multiple OPF formulations, including AC, DC, and second-order cone formulations. Standardized datasets are made publicly available to democratize access to this field, reduce the burden of data generation, and enable the fair comparison of various methodologies. PGLearn also includes a robust toolkit for training, evaluating, and benchmarking machine learning models for OPF, with the goal of standardizing performance evaluation across the field. By promoting open, standardized datasets and evaluation metrics, PGLearn aims at democratizing and accelerating research and innovation in machine learning applications for optimal power flow problems. Datasets are available for download at https://www.huggingface.co/PGLearn.
Authors: Chenruo Liu, Kenan Tang, Yao Qin, Qi Lei
Abstract: This paper bridges distribution shift and AI safety through a comprehensive analysis of their conceptual and methodological synergies. While prior discussions often focus on narrow cases or informal analogies, we establish two types connections between specific causes of distribution shift and fine-grained AI safety issues: (1) methods addressing a specific shift type can help achieve corresponding safety goals, or (2) certain shifts and safety issues can be formally reduced to each other, enabling mutual adaptation of their methods. Our findings provide a unified perspective that encourages fundamental integration between distribution shift and AI safety research.
Authors: Alexander Gill, Abhilasha Ravichander, Ana Marasovi\'c
Abstract: Large language models (LLMs) are increasingly used for data generation. However, creating evaluation benchmarks raises the bar for this emerging paradigm. Benchmarks must target specific phenomena, penalize exploiting shortcuts, and be challenging. Through two case studies, we investigate whether LLMs can meet these demands by generating reasoning over-text benchmarks and comparing them to those created through careful crowdsourcing. Specifically, we evaluate both the validity and difficulty of LLM-generated versions of two high-quality reading comprehension datasets: CondaQA, which evaluates reasoning about negation, and DROP, which targets reasoning about quantities. We find that prompting LLMs can produce variants of these datasets that are often valid according to the annotation guidelines, at a fraction of the cost of the original crowdsourcing effort. However, we show that they are less challenging for LLMs than their human-authored counterparts. This finding sheds light on what may have been lost by generating evaluation data with LLMs, and calls for critically reassessing the immediate use of this increasingly prevalent approach to benchmark creation.
Authors: Peiling Jiang, Haijun Xia
Abstract: Web-based activities are fundamentally distributed across webpages. However, conventional browsers with stacks of tabs fail to support operating and synthesizing large volumes of information across pages. While recent AI systems enable fully automated web browsing and information synthesis, they often diminish user agency and hinder contextual understanding. Therefore, we explore how AI could instead augment users' interactions with content across webpages and mitigate cognitive and manual efforts. Through literature on information tasks and web browsing challenges, and an iterative design process, we present a rich set of novel interactions with our prototype web browser, Orca. Leveraging AI, Orca supports user-driven exploration, operation, organization, and synthesis of web content at scale. To enable browsing at scale, webpages are treated as malleable materials that humans and AI can collaboratively manipulate and compose into a malleable, dynamic, and browser-level workspace. Our evaluation revealed an increased "appetite" for information foraging, enhanced user control, and more flexibility in sensemaking across a broader information landscape on the web.
Authors: Liu Yuezhang, Xue-Xin Wei
Abstract: Recent findings suggest that diffusion models significantly enhance empirical adversarial robustness. While some intuitive explanations have been proposed, the precise mechanisms underlying these improvements remain unclear. In this work, we systematically investigate how and how well diffusion models improve adversarial robustness. First, we observe that diffusion models intriguingly increase, rather than decrease, the $\ell_p$ distance to clean samples--challenging the intuition that purification denoises inputs closer to the original data. Second, we find that the purified images are heavily influenced by the internal randomness of diffusion models, where a compression effect arises within each randomness configuration. Motivated by this observation, we evaluate robustness under fixed randomness and find that the improvement drops to approximately 24% on CIFAR-10--substantially lower than prior reports approaching 70%. Importantly, we show that this remaining robustness gain strongly correlates with the model's ability to compress the input space, revealing the compression rate as a reliable robustness indicator without requiring gradient-based analysis. Our findings provide novel insights into the mechanisms underlying diffusion-based purification, and offer guidance for developing more effective and principled adversarial purification systems.
Authors: Alexander Herzog, Aliai Eusebi, Lorenzo Cavallaro
Abstract: The performance figures of modern drift-adaptive malware classifiers appear promising, but does this translate to genuine operational reliability? The standard evaluation paradigm primarily focuses on baseline performance metrics, neglecting confidence-error alignment and operational stability. While TESSERACT established the importance of temporal evaluation, we take a complementary direction by investigating whether malware classifiers maintain reliable confidence estimates under distribution shifts and exploring the tensions between scientific advancement and practical impacts when they do not. We propose AURORA, a framework to evaluate malware classifiers based on their confidence quality and operational resilience. AURORA subjects the confidence profile of a given model to verification to assess the reliability of its estimates. Unreliable confidence estimates erode operational trust, waste valuable annotation budget on non-informative samples for active learning, and leave error-prone instances undetected in selective classification. AURORA is further complemented by a set of metrics designed to go beyond point-in-time performance, striving towards a more holistic assessment of operational stability throughout temporal evaluation periods. The fragility we observe in state-of-the-art frameworks across datasets of varying drift severity suggests the need for a return to the whiteboard.
Authors: Sandra H\"oltervennhoff, Jonas Ricker, Maike M. Raphael, Charlotte Schwedes, Rebecca Weil, Asja Fischer, Thorsten Holz, Lea Sch\"onherr, Sascha Fahl
Abstract: Generative artificial intelligence is developing rapidly, impacting humans' interaction with information and digital media. It is increasingly used to create deceptively realistic misinformation, so lawmakers have imposed regulations requiring the disclosure of AI-generated content. However, only little is known about whether these labels reduce the risks of AI-generated misinformation. Our work addresses this research gap. Focusing on AI-generated images, we study the implications of labels, including the possibility of mislabeling. Assuming that simplicity, transparency, and trust are likely to impact the successful adoption of such labels, we first qualitatively explore users' opinions and expectations of AI labeling using five focus groups. Second, we conduct a pre-registered online survey with over 1300 U.S. and EU participants to quantitatively assess the effect of AI labels on users' ability to recognize misinformation containing either human-made or AI-generated images. Our focus groups illustrate that, while participants have concerns about the practical implementation of labeling, they consider it helpful in identifying AI-generated images and avoiding deception. However, considering security benefits, our survey revealed an ambiguous picture, suggesting that users might over-rely on labels. While inaccurate claims supported by labeled AI-generated images were rated less credible than those with unlabeled AI-images, the belief in accurate claims also decreased when accompanied by a labeled AI-generated image. Moreover, we find the undesired side effect that human-made images conveying inaccurate claims were perceived as more credible in the presence of labels.
Authors: Nikita Khramov, Andrei Kozyrev, Gleb Solovev, Anton Podkopaev
Abstract: Interactive Theorem Proving was repeatedly shown to be fruitful combined with Generative Artificial Intelligence. This paper assesses multiple approaches to Rocq generation and illuminates potential avenues for improvement. We highlight the importance of thorough premise selection for generating Rocq proofs and propose a novel approach, leveraging retrieval via a self-attentive embedder model. The evaluation of the designed approach shows up to 28% relative increase of the generator's performance. We tackle the problem of writing Rocq proofs using a multi-stage agentic system, tailored for formal verification, and demonstrate its high effectiveness. We conduct an ablation study and show the use of multi-agent debate on the planning stage of proof synthesis.
Authors: Krti Tallam, Emma Miller
Abstract: CaMeL (Capabilities for Machine Learning) introduces a capability-based sandbox to mitigate prompt injection attacks in large language model (LLM) agents. While effective, CaMeL assumes a trusted user prompt, omits side-channel concerns, and incurs performance tradeoffs due to its dual-LLM design. This response identifies these issues and proposes engineering improvements to expand CaMeL's threat coverage and operational usability. We introduce: (1) prompt screening for initial inputs, (2) output auditing to detect instruction leakage, (3) a tiered-risk access model to balance usability and control, and (4) a verified intermediate language for formal guarantees. Together, these upgrades align CaMeL with best practices in enterprise security and support scalable deployment.
Authors: Vladimir Bataev, Andrei Andrusenko, Lilit Grigoryan, Aleksandr Laptev, Vitaly Lavrukhin, Boris Ginsburg
Abstract: Statistical n-gram language models are widely used for context-biasing tasks in Automatic Speech Recognition (ASR). However, existing implementations lack computational efficiency due to poor parallelization, making context-biasing less appealing for industrial use. This work rethinks data structures for statistical n-gram language models to enable fast and parallel operations for GPU-optimized inference. Our approach, named NGPU-LM, introduces customizable greedy decoding for all major ASR model types - including transducers, attention encoder-decoder models, and CTC - with less than 7% computational overhead. The proposed approach can eliminate more than 50% of the accuracy gap between greedy and beam search for out-of-domain scenarios while avoiding significant slowdown caused by beam search. The implementation of the proposed NGPU-LM is open-sourced.
Authors: Bargav Jayaraman, Virendra J. Marathe, Hamid Mozaffari, William F. Shen, Krishnaram Kenthapadi
Abstract: In enterprise settings, organizational data is segregated, siloed and carefully protected by elaborate access control frameworks. These access control structures can completely break down if an LLM fine-tuned on the siloed data serves requests, for downstream tasks, from individuals with disparate access privileges. We propose Permissioned LLMs (PermLLM), a new class of LLMs that superimpose the organizational data access control structures on query responses they generate. We formalize abstractions underpinning the means to determine whether access control enforcement happens correctly over LLM query responses. Our formalism introduces the notion of a relevant response that can be used to prove whether a PermLLM mechanism has been implemented correctly. We also introduce a novel metric, called access advantage, to empirically evaluate the efficacy of a PermLLM mechanism. We introduce three novel PermLLM mechanisms that build on Parameter Efficient Fine-Tuning to achieve the desired access control. We furthermore present two instantiations of access advantage--(i) Domain Distinguishability Index (DDI) based on Membership Inference Attacks, and (ii) Utility Gap Index (UGI) based on LLM utility evaluation. We demonstrate the efficacy of our PermLLM mechanisms through extensive experiments on four public datasets (GPQA, RCV1, SimpleQA, and WMDP), in addition to evaluating the validity of DDI and UGI metrics themselves for quantifying access control in LLMs.
Authors: Susan Liang, Dejan Markovic, Israel D. Gebru, Steven Krenn, Todd Keebler, Jacob Sandakly, Frank Yu, Samuel Hassel, Chenliang Xu, Alexander Richard
Abstract: Binaural rendering aims to synthesize binaural audio that mimics natural hearing based on a mono audio and the locations of the speaker and listener. Although many methods have been proposed to solve this problem, they struggle with rendering quality and streamable inference. Synthesizing high-quality binaural audio that is indistinguishable from real-world recordings requires precise modeling of binaural cues, room reverb, and ambient sounds. Additionally, real-world applications demand streaming inference. To address these challenges, we propose a flow matching based streaming binaural speech synthesis framework called BinauralFlow. We consider binaural rendering to be a generation problem rather than a regression problem and design a conditional flow matching model to render high-quality audio. Moreover, we design a causal U-Net architecture that estimates the current audio frame solely based on past information to tailor generative models for streaming inference. Finally, we introduce a continuous inference pipeline incorporating streaming STFT/ISTFT operations, a buffer bank, a midpoint solver, and an early skip schedule to improve rendering continuity and speed. Quantitative and qualitative evaluations demonstrate the superiority of our method over SOTA approaches. A perceptual study further reveals that our model is nearly indistinguishable from real-world recordings, with a $42\%$ confusion rate.
Authors: Nicolas Espinosa-Dice, Yiyi Zhang, Yiding Chen, Bradley Guo, Owen Oertell, Gokul Swamy, Kiante Brantley, Wen Sun
Abstract: Diffusion and flow models have emerged as powerful generative approaches capable of modeling diverse and multimodal behavior. However, applying these models to offline reinforcement learning (RL) remains challenging due to the iterative nature of their noise sampling processes, making policy optimization difficult. In this paper, we introduce Scalable Offline Reinforcement Learning (SORL), a new offline RL algorithm that leverages shortcut models - a novel class of generative models - to scale both training and inference. SORL's policy can capture complex data distributions and can be trained simply and efficiently in a one-stage training procedure. At test time, SORL introduces both sequential and parallel inference scaling by using the learned Q-function as a verifier. We demonstrate that SORL achieves strong performance across a range of offline RL tasks and exhibits positive scaling behavior with increased test-time compute. We release the code at nico-espinosadice.github.io/projects/sorl.
Authors: Shams Tarek, Dipayan Saha, Sujan Kumar Saha, Farimah Farahmandi
Abstract: The current landscape of system-on-chips (SoCs) security verification faces challenges due to manual, labor-intensive, and inflexible methodologies. These issues limit the scalability and effectiveness of security protocols, making bug detection at the Register-Transfer Level (RTL) difficult. This paper proposes a new framework named BugWhisperer that utilizes a specialized, fine-tuned Large Language Model (LLM) to address these challenges. By enhancing the LLM's hardware security knowledge and leveraging its capabilities for text inference and knowledge transfer, this approach automates and improves the adaptability and reusability of the verification process. We introduce an open-source, fine-tuned LLM specifically designed for detecting security vulnerabilities in SoC designs. Our findings demonstrate that this tailored LLM effectively enhances the efficiency and flexibility of the security verification process. Additionally, we introduce a comprehensive hardware vulnerability database that supports this work and will further assist the research community in enhancing the security verification process.
Authors: Xiaoyang Zhan, Shixin Zhou, Qianqian Yang, Yixuan Zhao, Hao Liu, Srinivas Chowdary Ramineni, Kenji Shimada
Abstract: This paper presents a system for autonomous semantic exploration and dense semantic target mapping of a complex unknown environment using a ground robot equipped with a LiDAR-panoramic camera suite. Existing approaches often struggle to balance collecting high-quality observations from multiple view angles and avoiding unnecessary repetitive traversal. To fill this gap, we propose a complete system combining mapping and planning. We first redefine the task as completing both geometric coverage and semantic viewpoint observation. We then manage semantic and geometric viewpoints separately and propose a novel Priority-driven Decoupled Local Sampler to generate local viewpoint sets. This enables explicit multi-view semantic inspection and voxel coverage without unnecessary repetition. Building on this, we develop a hierarchical planner to ensure efficient global coverage. In addition, we propose a Safe Aggressive Exploration State Machine, which allows aggressive exploration behavior while ensuring the robot's safety. Our system includes a plug-and-play semantic target mapping module that integrates seamlessly with state-of-the-art SLAM algorithms for pointcloud-level dense semantic target mapping. We validate our approach through extensive experiments in both realistic simulations and complex real-world environments. Simulation results show that our planner achieves faster exploration and shorter travel distances while guaranteeing a specified number of multi-view inspections. Real-world experiments further confirm the system's effectiveness in achieving accurate dense semantic object mapping of unstructured environments.
Authors: Hamidreza Montazeri Hedesh, Moh Kamalul Wafi, Milad Siami
Abstract: We study the local stability of nonlinear systems in the Lur'e form with static nonlinear feedback realized by feedforward neural networks (FFNNs). By leveraging positivity system constraints, we employ a localized variant of the Aizerman conjecture, which provides sufficient conditions for exponential stability of trajectories confined to a compact set. Using this foundation, we develop two distinct methods for estimating the Region of Attraction (ROA): (i) a less conservative Lyapunov-based approach that constructs invariant sublevel sets of a quadratic function satisfying a linear matrix inequality (LMI), and (ii) a novel technique for computing tight local sector bounds for FFNNs via layer-wise propagation of linear relaxations. These bounds are integrated into the localized Aizerman framework to certify local exponential stability. Numerical results demonstrate substantial improvements over existing integral quadratic constraint-based approaches in both ROA size and scalability.
Authors: Youngsoo Choi, Siu Wun Cheung, Youngkyu Kim, Ping-Hsuan Tsai, Alejandro N. Diaz, Ivan Zanardi, Seung Whan Chung, Dylan Matthew Copeland, Coleman Kendrick, William Anderson, Traian Iliescu, Matthias Heinkenschloss
Abstract: The widespread success of foundation models in natural language processing and computer vision has inspired researchers to extend the concept to scientific machine learning and computational science. However, this position paper argues that as the term "foundation model" is an evolving concept, its application in computational science is increasingly used without a universally accepted definition, potentially creating confusion and diluting its precise scientific meaning. In this paper, we address this gap by proposing a formal definition of foundation models in computational science, grounded in the core values of generality, reusability, and scalability. We articulate a set of essential and desirable characteristics that such models must exhibit, drawing parallels with traditional foundational methods, like the finite element and finite volume methods. Furthermore, we introduce the Data-Driven Finite Element Method (DD-FEM), a framework that fuses the modular structure of classical FEM with the representational power of data-driven learning. We demonstrate how DD-FEM addresses many of the key challenges in realizing foundation models for computational science, including scalability, adaptability, and physics consistency. By bridging traditional numerical methods with modern AI paradigms, this work provides a rigorous foundation for evaluating and developing novel approaches toward future foundation models in computational science.
Authors: Emmanuel Anaya Gonz\'alez, Raven Rothkopf, Sorin Lerner, Nadia Polikarpova
Abstract: While AI programming tools hold the promise of increasing programmers' capabilities and productivity to a remarkable degree, they often exclude users from essential decision-making processes, causing many to effectively "turn off their brains" and over-rely on solutions provided by these systems. These behaviors can have severe consequences in critical domains, like software security. We propose Human-in-the-loop Decoding, a novel interaction technique that allows users to observe and directly influence LLM decisions during code generation, in order to align the model's output with their personal requirements. We implement this technique in HiLDe, a code completion assistant that highlights critical decisions made by the LLM and provides local alternatives for the user to explore. In a within-subjects study (N=18) on security-related tasks, we found that HiLDe led participants to generate significantly fewer vulnerabilities and better align code generation with their goals compared to a traditional code completion assistant.
Authors: Cristian Chica, Yinglong Guo, Gilad Lerman
Abstract: There is growing experimental evidence that $Q$-learning agents may learn to charge supracompetitive prices. We provide the first theoretical explanation for this behavior in infinite repeated games. Firms update their pricing policies based solely on observed profits, without computing equilibrium strategies. We show that when the game admits both a one-stage Nash equilibrium price and a collusive-enabling price, and when the $Q$-function satisfies certain inequalities at the end of experimentation, firms learn to consistently charge supracompetitive prices. We introduce a new class of one-memory subgame perfect equilibria (SPEs) and provide conditions under which learned behavior is supported by naive collusion, grim trigger policies, or increasing strategies. Naive collusion does not constitute an SPE unless the collusive-enabling price is a one-stage Nash equilibrium, whereas grim trigger policies can.
Authors: Athanasios Glentis, Jiaxiang Li, Qiulin Shang, Andi Han, Ioannis Tsaknakis, Quan Wei, Mingyi Hong
Abstract: Fueled by their remarkable ability to tackle diverse tasks across multiple domains, large language models (LLMs) have grown at an unprecedented rate, with some recent models containing trillions of parameters. This growth is accompanied by substantial computational challenges, particularly regarding the memory and compute resources required for training and fine-tuning. Numerous approaches have been explored to address these issues, such as LoRA. While these methods are effective for fine-tuning, their application to pre-training is significantly more challenging due to the need to learn vast datasets. Motivated by this issue, we aim to address the following questions: Can parameter- or memory-efficient methods enhance pre-training efficiency while achieving performance comparable to full-model training? How can the performance gap be narrowed? To this end, the contributions of this work are the following. (1) We begin by conducting a comprehensive survey that summarizes state-of-the-art methods for efficient pre-training. (2) We perform a benchmark evaluation of several representative memory efficient pre-training approaches to comprehensively evaluate their performance across model sizes. We observe that with a proper choice of optimizer and hyperparameters, full-rank training delivers the best performance, as expected. We also notice that incorporating high-rank updates in low-rank approaches is the key to improving their performance. (3) Finally, we propose two practical techniques, namely weight refactorization and momentum reset, to enhance the performance of efficient pre-training methods. We observe that applying these techniques to the low-rank method (on a 1B model) can achieve a lower perplexity than popular memory efficient algorithms such as GaLore and Fira, while simultaneously using about 25% less memory.
Authors: Haobo Zhang, Jiayu Zhou
Abstract: Fine-tuning large language models (LMs) for individual tasks yields strong performance but is expensive for deployment and storage. Recent works explore model merging to combine multiple task-specific models into a single multi-task model without additional training. However, existing merging methods often fail for models fine-tuned with low-rank adaptation (LoRA), due to significant performance degradation. In this paper, we show that this issue arises from a previously overlooked interplay between model parameters and data distributions. We propose Orthogonal Subspaces for Robust model Merging (OSRM) to constrain the LoRA subspace *prior* to fine-tuning, ensuring that updates relevant to one task do not adversely shift outputs for others. Our approach can seamlessly integrate with most existing merging algorithms, reducing the unintended interference among tasks. Extensive experiments on eight datasets, tested with three widely used LMs and two large LMs, demonstrate that our method not only boosts merging performance but also preserves single-task accuracy. Furthermore, our approach exhibits greater robustness to the hyperparameters of merging. These results highlight the importance of data-parameter interaction in model merging and offer a plug-and-play solution for merging LoRA models.
Authors: Niclas Boehmer, Sara Fish, Ariel D. Procaccia
Abstract: A key task in certain democratic processes is to produce a concise slate of statements that proportionally represents the full spectrum of user opinions. This task is similar to committee elections, but unlike traditional settings, the candidate set comprises all possible statements of varying lengths, and so it can only be accessed through specific queries. Combining social choice and large language models, prior work has approached this challenge through a framework of generative social choice. We extend the framework in two fundamental ways, providing theoretical guarantees even in the face of approximately optimal queries and a budget limit on the overall length of the slate. Using GPT-4o to implement queries, we showcase our approach on datasets related to city improvement measures and drug reviews, demonstrating its effectiveness in generating representative slates from unstructured user opinions.
Authors: Yuchen Zhuang, Di Jin, Jiaao Chen, Wenqi Shi, Hanrui Wang, Chao Zhang
Abstract: Large language models (LLMs)-empowered web agents enables automating complex, real-time web navigation tasks in enterprise environments. However, existing web agents relying on supervised fine-tuning (SFT) often struggle with generalization and robustness due to insufficient reasoning capabilities when handling the inherently dynamic nature of web interactions. In this study, we introduce WorkForceAgent-R1, an LLM-based web agent trained using a rule-based R1-style reinforcement learning framework designed explicitly to enhance single-step reasoning and planning for business-oriented web navigation tasks. We employ a structured reward function that evaluates both adherence to output formats and correctness of actions, enabling WorkForceAgent-R1 to implicitly learn robust intermediate reasoning without explicit annotations or extensive expert demonstrations. Extensive experiments on the WorkArena benchmark demonstrate that WorkForceAgent-R1 substantially outperforms SFT baselines by 10.26-16.59%, achieving competitive performance relative to proprietary LLM-based agents (gpt-4o) in workplace-oriented web navigation tasks.
Authors: Jaewoo Ahn, Heeseung Yun, Dayoon Ko, Gunhee Kim
Abstract: While pre-trained multimodal representations (e.g., CLIP) have shown impressive capabilities, they exhibit significant compositional vulnerabilities leading to counterintuitive judgments. We introduce Multimodal Adversarial Compositionality (MAC), a benchmark that leverages large language models (LLMs) to generate deceptive text samples to exploit these vulnerabilities across different modalities and evaluates them through both sample-wise attack success rate and group-wise entropy-based diversity. To improve zero-shot methods, we propose a self-training approach that leverages rejection-sampling fine-tuning with diversity-promoting filtering, which enhances both attack success rate and sample diversity. Using smaller language models like Llama-3.1-8B, our approach demonstrates superior performance in revealing compositional vulnerabilities across various multimodal representations, including images, videos, and audios.
Authors: Angtian Wang, Haibin Huang, Jacob Zhiyuan Fang, Yiding Yang, Chongyang Ma
Abstract: We propose a unified framework for motion control in video generation that seamlessly integrates camera movement, object-level translation, and fine-grained local motion using trajectory-based inputs. In contrast to prior methods that address these motion types through separate modules or task-specific designs, our approach offers a cohesive solution by projecting user-defined trajectories into the latent space of pre-trained image-to-video generation models via a lightweight motion injector. Users can specify keypoints and their motion paths to control localized deformations, entire object motion, virtual camera dynamics, or combinations of these. The injected trajectory signals guide the generative process to produce temporally consistent and semantically aligned motion sequences. Our framework demonstrates superior performance across multiple video motion control tasks, including stylized motion effects (e.g., motion brushes), dynamic viewpoint changes, and precise local motion manipulation. Experiments show that our method provides significantly better controllability and visual quality compared to prior approaches and commercial solutions, while remaining broadly compatible with various state-of-the-art video generation backbones. Project page: https://anytraj.github.io/.
Authors: Alisha Srivastava, Emir Korukluoglu, Minh Nhat Le, Duyen Tran, Chau Minh Pham, Marzena Karpinska, Mohit Iyyer
Abstract: Large language models (LLMs) are known to memorize and recall English text from their pretraining data. However, the extent to which this ability generalizes to non-English languages or transfers across languages remains unclear. This paper investigates multilingual and cross-lingual memorization in LLMs, probing if memorized content in one language (e.g., English) can be recalled when presented in translation. To do so, we introduce OWL, a dataset of 31.5K aligned excerpts from 20 books in ten languages, including English originals, official translations (Vietnamese, Spanish, Turkish), and new translations in six low-resource languages (Sesotho, Yoruba, Maithili, Malagasy, Setswana, Tahitian). We evaluate memorization across model families and sizes through three tasks: (1) direct probing, which asks the model to identify a book's title and author; (2) name cloze, which requires predicting masked character names; and (3) prefix probing, which involves generating continuations. We find that LLMs consistently recall content across languages, even for texts without direct translation in pretraining data. GPT-4o, for example, identifies authors and titles 69% of the time and masked entities 6% of the time in newly translated excerpts. Perturbations (e.g., masking characters, shuffling words) modestly reduce direct probing accuracy (7% drop for shuffled official translations). Our results highlight the extent of cross-lingual memorization and provide insights on the differences between the models.
Authors: Yuhui Zhang, Yuchang Su, Yiming Liu, Serena Yeung-Levy
Abstract: Negation is a fundamental linguistic phenomenon that can entirely reverse the meaning of a sentence. As vision language models (VLMs) continue to advance and are deployed in high-stakes applications, assessing their ability to comprehend negation becomes essential. To address this, we introduce NegVQA, a visual question answering (VQA) benchmark consisting of 7,379 two-choice questions covering diverse negation scenarios and image-question distributions. We construct NegVQA by leveraging large language models to generate negated versions of questions from existing VQA datasets. Evaluating 20 state-of-the-art VLMs across seven model families, we find that these models struggle significantly with negation, exhibiting a substantial performance drop compared to their responses to the original questions. Furthermore, we uncover a U-shaped scaling trend, where increasing model size initially degrades performance on NegVQA before leading to improvements. Our benchmark reveals critical gaps in VLMs' negation understanding and offers insights into future VLM development. Project page available at https://yuhui-zh15.github.io/NegVQA/.
Authors: Sheng Zhang, Qin Liu, Naoto Usuyama, Cliff Wong, Tristan Naumann, Hoifung Poon
Abstract: The emergence of scaling laws has profoundly shaped the development of large language models (LLMs), enabling predictable performance gains through systematic increases in model size, dataset volume, and compute. Yet, these principles remain largely unexplored in the context of electronic health records (EHRs) -- a rich, sequential, and globally abundant data source that differs structurally from natural language. In this work, we present the first empirical investigation of scaling laws for EHR foundation models. By training transformer architectures on patient timeline data from the MIMIC-IV database across varying model sizes and compute budgets, we identify consistent scaling patterns, including parabolic IsoFLOPs curves and power-law relationships between compute, model parameters, data size, and clinical utility. These findings demonstrate that EHR models exhibit scaling behavior analogous to LLMs, offering predictive insights into resource-efficient training strategies. Our results lay the groundwork for developing powerful EHR foundation models capable of transforming clinical prediction tasks and advancing personalized healthcare.
Authors: Bahareh Tolooshams, Aditi Chandrashekar, Rayhan Zirvi, Abbas Mammadov, Jiachen Yao, Chuwei Wang, Anima Anandkumar
Abstract: Diffusion models represent the state-of-the-art for solving inverse problems such as image restoration tasks. In the Bayesian framework, diffusion-based inverse solvers incorporate a likelihood term to guide the prior sampling process, generating data consistent with the posterior distribution. However, due to the intractability of the likelihood term, many current methods rely on isotropic Gaussian approximations, which lead to deviations from the data manifold and result in inconsistent, unstable reconstructions. We propose Equivariance Regularized (EquiReg) diffusion, a general framework for regularizing posterior sampling in diffusion-based inverse problem solvers. EquiReg enhances reconstructions by reweighting diffusion trajectories and penalizing those that deviate from the data manifold. We define a new distribution-dependent equivariance error, empirically identify functions that exhibit low error for on-manifold samples and higher error for off-manifold samples, and leverage these functions to regularize the diffusion sampling process. When applied to a variety of solvers, EquiReg outperforms state-of-the-art diffusion models in both linear and nonlinear image restoration tasks, as well as in reconstructing partial differential equations.
Authors: Kewei Lian, Shaofei Cai, Yilun Du, Yitao Liang
Abstract: The ability to simulate the world in a spatially consistent manner is a crucial requirements for effective world models. Such a model enables high-quality visual generation, and also ensures the reliability of world models for downstream tasks such as simulation and planning. Designing a memory module is a crucial component for addressing spatial consistency: such a model must not only retain long-horizon observational information, but also enables the construction of explicit or implicit internal spatial representations. However, there are no dataset designed to promote the development of memory modules by explicitly enforcing spatial consistency constraints. Furthermore, most existing benchmarks primarily emphasize visual coherence or generation quality, neglecting the requirement of long-range spatial consistency. To bridge this gap, we construct a dataset and corresponding benchmark by sampling 150 distinct locations within the open-world environment of Minecraft, collecting about 250 hours (20 million frames) of loop-based navigation videos with actions. Our dataset follows a curriculum design of sequence lengths, allowing models to learn spatial consistency on increasingly complex navigation trajectories. Furthermore, our data collection pipeline is easily extensible to new Minecraft environments and modules. Four representative world model baselines are evaluated on our benchmark. Dataset, benchmark, and code are open-sourced to support future research.
Authors: Masaharu Kagiyama, Tsuyoshi Okita
Abstract: This paper aims to develop an energy-efficient classifier for time-series data by introducing PatchEchoClassifier, a novel model that leverages a reservoir-based mechanism known as the Echo State Network (ESN). The model is designed for human activity recognition (HAR) using one-dimensional sensor signals and incorporates a tokenizer to extract patch-level representations. To train the model efficiently, we propose a knowledge distillation framework that transfers knowledge from a high-capacity MLP-Mixer teacher to the lightweight reservoir-based student model. Experimental evaluations on multiple HAR datasets demonstrate that our model achieves over 80 percent accuracy while significantly reducing computational cost. Notably, PatchEchoClassifier requires only about one-sixth of the floating point operations (FLOPS) compared to DeepConvLSTM, a widely used convolutional baseline. These results suggest that PatchEchoClassifier is a promising solution for real-time and energy-efficient human activity recognition in edge computing environments.
Authors: Albert Tseng, Zhaofeng Sun, Christopher De Sa
Abstract: The main goal of post-training quantization (PTQ) is to produced a compressed model whose output distribution is as close to the original model's as possible. To do this tractably, almost all LLM PTQ algorithms quantize linear layers by independently minimizing the immediate activation error. However, this localized objective ignores the effect of subsequent layers, so reducing it does not necessarily give a closer model. In this work, we introduce Yet Another Quantization Algorithm (YAQA), an adaptive rounding algorithm that uses Kronecker-factored approximations of each linear layer's Hessian with respect to the \textit{full model} KL divergence. YAQA consists of two components: Kronecker-factored sketches of the full layerwise Hessian that can be tractably computed for hundred-billion parameter LLMs, and a quantizer-independent rounding algorithm that uses these sketches and comes with theoretical guarantees. Across a wide range of models and quantizers, YAQA empirically reduces the KL divergence to the original model by $\approx 30\%$ while achieving state of the art performance on downstream tasks.
Authors: Hoang Pham, Thanh-Do Nguyen, Khac-Hoai Nam Bui
Abstract: Claim verification is a long-standing and challenging task that demands not only high accuracy but also explainability of the verification process. This task becomes an emerging research issue in the era of large language models (LLMs) since real-world claims are often complex, featuring intricate semantic structures or obfuscated entities. Traditional approaches typically address this by decomposing claims into sub-claims and querying a knowledge base to resolve hidden or ambiguous entities. However, the absence of effective disambiguation strategies for these entities can compromise the entire verification process. To address these challenges, we propose Verify-in-the-Graph (VeGraph), a novel framework leveraging the reasoning and comprehension abilities of LLM agents. VeGraph operates in three phases: (1) Graph Representation - an input claim is decomposed into structured triplets, forming a graph-based representation that integrates both structured and unstructured information; (2) Entity Disambiguation -VeGraph iteratively interacts with the knowledge base to resolve ambiguous entities within the graph for deeper sub-claim verification; and (3) Verification - remaining triplets are verified to complete the fact-checking process. Experiments using Meta-Llama-3-70B (instruct version) show that VeGraph achieves competitive performance compared to baselines on two benchmarks HoVer and FEVEROUS, effectively addressing claim verification challenges. Our source code and data are available for further exploitation.
Authors: Linh Le Pham Van, Minh Hoang Nguyen, Hung Le, Hung The Tran, Sunil Gupta
Abstract: Robust reinforcement learning (RL) aims to learn policies that remain effective despite uncertainties in its environment, which frequently arise in real-world applications due to variations in environment dynamics. The robust RL methods learn a robust policy by maximizing value under the worst-case models within a predefined uncertainty set. Offline robust RL algorithms are particularly promising in scenarios where only a fixed dataset is available and new data cannot be collected. However, these approaches often require extensive offline data, and gathering such datasets for specific tasks in specific environments can be both costly and time-consuming. Using an imperfect simulator offers a faster, cheaper, and safer way to collect data for training, but it can suffer from dynamics mismatch. In this paper, we introduce HYDRO, the first Hybrid Cross-Domain Robust RL framework designed to address these challenges. HYDRO utilizes an online simulator to complement the limited amount of offline datasets in the non-trivial context of robust RL. By measuring and minimizing performance gaps between the simulator and the worst-case models in the uncertainty set, HYDRO employs novel uncertainty filtering and prioritized sampling to select the most relevant and reliable simulator samples. Our extensive experiments demonstrate HYDRO's superior performance over existing methods across various tasks, underscoring its potential to improve sample efficiency in offline robust RL.
Authors: Chiwan Park, Wonjun Jang, Daeryong Kim, Aelim Ahn, Kichang Yang, Woosung Hwang, Jihyeon Roh, Hyerin Park, Hyosun Wang, Min Seok Kim, Jihoon Kang
Abstract: The advancement of Large Language Models (LLMs) has led to significant improvements in various service domains, including search, recommendation, and chatbot applications. However, applying state-of-the-art (SOTA) research to industrial settings presents challenges, as it requires maintaining flexible conversational abilities while also strictly complying with service-specific constraints. This can be seen as two conflicting requirements due to the probabilistic nature of LLMs. In this paper, we propose our approach to addressing this challenge and detail the strategies we employed to overcome their inherent limitations in real-world applications. We conduct a practical case study of a conversational agent designed for the e-commerce domain, detailing our implementation workflow and optimizations. Our findings provide insights into bridging the gap between academic research and real-world application, introducing a framework for developing scalable, controllable, and reliable AI-driven agents.
Authors: Jonathan Li, Zoltan Csaki, Nidhi Hiremath, Etash Guha, Fenglu Hong, Edward Ma, Urmish Thakker
Abstract: Modern VLMs have achieved near-saturation accuracy in English document visual question-answering (VQA). However, this task remains challenging in lower resource languages due to a dearth of suitable training and evaluation data. In this paper we present scalable methods for curating such datasets by focusing on Hungarian, approximately the 17th highest resource language on the internet. Specifically, we present HuDocVQA and HuDocVQA-manual, document VQA datasets that modern VLMs significantly underperform on compared to English DocVQA. HuDocVQA-manual is a small manually curated dataset based on Hungarian documents from Common Crawl, while HuDocVQA is a larger synthetically generated VQA data set from the same source. We apply multiple rounds of quality filtering and deduplication to HuDocVQA in order to match human-level quality in this dataset. We also present HuCCPDF, a dataset of 117k pages from Hungarian Common Crawl PDFs along with their transcriptions, which can be used for training a model for Hungarian OCR. To validate the quality of our datasets, we show how finetuning on a mixture of these datasets can improve accuracy on HuDocVQA for Llama 3.2 11B Instruct by +7.2%. Our datasets and code will be released to the public to foster further research in multilingual DocVQA.
Authors: Xingjian Wu, Xiangfei Qiu, Hongfan Gao, Jilin Hu, Bin Yang, Chenjuan Guo
Abstract: Probabilistic Time Series Forecasting (PTSF) plays a crucial role in decision-making across various fields, including economics, energy, and transportation. Most existing methods excell at short-term forecasting, while overlooking the hurdles of Long-term Probabilistic Time Series Forecasting (LPTSF). As the forecast horizon extends, the inherent nonlinear dynamics have a significant adverse effect on prediction accuracy, and make generative models inefficient by increasing the cost of each iteration. To overcome these limitations, we introduce $K^2$VAE, an efficient VAE-based generative model that leverages a KoopmanNet to transform nonlinear time series into a linear dynamical system, and devises a KalmanNet to refine predictions and model uncertainty in such linear system, which reduces error accumulation in long-term forecasting. Extensive experiments demonstrate that $K^2$VAE outperforms state-of-the-art methods in both short- and long-term PTSF, providing a more efficient and accurate solution.
Authors: Jinchuan Zhang, Lu Yin, Yan Zhou, Songlin Hu
Abstract: The acquisition of agentic capabilities has transformed LLMs from "knowledge providers" to "action executors", a trend that while expanding LLMs' capability boundaries, significantly increases their susceptibility to malicious use. Previous work has shown that current LLM-based agents execute numerous malicious tasks even without being attacked, indicating a deficiency in agentic use safety alignment during the post-training phase. To address this gap, we propose AgentAlign, a novel framework that leverages abstract behavior chains as a medium for safety alignment data synthesis. By instantiating these behavior chains in simulated environments with diverse tool instances, our framework enables the generation of highly authentic and executable instructions while capturing complex multi-step dynamics. The framework further ensures model utility by proportionally synthesizing benign instructions through non-malicious interpretations of behavior chains, precisely calibrating the boundary between helpfulness and harmlessness. Evaluation results on AgentHarm demonstrate that fine-tuning three families of open-source models using our method substantially improves their safety (35.8% to 79.5% improvement) while minimally impacting or even positively enhancing their helpfulness, outperforming various prompting methods. The dataset and code have both been open-sourced.
Authors: Haewon Park, Gyubin Choi, Minjun Kim, Yohan Jo
Abstract: Knowledge editing (KE) methods offer an efficient way to modify knowledge in large language models. Current KE evaluations typically assess editing success by considering only the edited knowledge without any preceding contexts. In real-world applications, however, preceding contexts often trigger the retrieval of the original knowledge and undermine the intended edit. To address this issue, we develop CHED -- a benchmark designed to evaluate the context robustness of KE methods. Evaluations on CHED show that they often fail when preceding contexts are present. To mitigate this shortcoming, we introduce CoRE, a KE method designed to strengthen context robustness by minimizing context-sensitive variance in hidden states of the model for edited knowledge. This method not only improves the editing success rate in situations where a preceding context is present but also preserves the overall capabilities of the model. We provide an in-depth analysis of the differing impacts of preceding contexts when introduced as user utterances versus assistant responses, and we dissect attention-score patterns to assess how specific tokens influence editing success.
Authors: Minh Nguyen Nhat To, Paul F RWilson, Viet Nguyen, Mohamed Harmanani, Michael Cooper, Fahimeh Fooladgar, Purang Abolmaesumi, Parvin Mousavi, Rahul G. Krishnan
Abstract: The subpopulationtion shift, characterized by a disparity in subpopulation distributibetween theween the training and target datasets, can significantly degrade the performance of machine learning models. Current solutions to subpopulation shift involve modifying empirical risk minimization with re-weighting strategies to improve generalization. This strategy relies on assumptions about the number and nature of subpopulations and annotations on group membership, which are unavailable for many real-world datasets. Instead, we propose using an ensemble of diverse classifiers to adaptively capture risk associated with subpopulations. Given a feature extractor network, we replace its standard linear classification layer with a mixture of prototypical classifiers, where each member is trained to classify the data while focusing on different features and samples from other members. In empirical evaluation on nine real-world datasets, covering diverse domains and kinds of subpopulation shift, our method of Diverse Prototypical Ensembles (DPEs) often outperforms the prior state-of-the-art in worst-group accuracy. The code is available at https://github.com/minhto2802/dpe4subpop
Authors: Dongwoo Lee, Dong Bok Lee, Steven Adriaensen, Juho Lee, Sung Ju Hwang, Frank Hutter, Seon Joo Kim, Hae Beom Lee
Abstract: Scaling has been a major driver of recent advancements in deep learning. Numerous empirical studies have found that scaling laws often follow the power-law and proposed several variants of power-law functions to predict the scaling behavior at larger scales. However, existing methods mostly rely on point estimation and do not quantify uncertainty, which is crucial for real-world applications involving decision-making problems such as determining the expected performance improvements achievable by investing additional computational resources. In this work, we explore a Bayesian framework based on Prior-data Fitted Networks (PFNs) for neural scaling law extrapolation. Specifically, we design a prior distribution that enables the sampling of infinitely many synthetic functions resembling real-world neural scaling laws, allowing our PFN to meta-learn the extrapolation. We validate the effectiveness of our approach on real-world neural scaling laws, comparing it against both the existing point estimation methods and Bayesian approaches. Our method demonstrates superior performance, particularly in data-limited scenarios such as Bayesian active learning, underscoring its potential for reliable, uncertainty-aware extrapolation in practical applications.
Authors: Siwen Wang, Shitou Zhang, Wan-Lin Chen, Dung Truong, Tzyy-Ping Jung
Abstract: Recent advancements in Large Language Models have inspired the development of foundation models across various domains. In this study, we evaluate the efficacy of Large EEG Models (LEMs) by fine-tuning LaBraM, a state-of-the-art foundation EEG model, on a real-world stress classification dataset collected in a graduate classroom. Unlike previous studies that primarily evaluate LEMs using data from controlled clinical settings, our work assesses their applicability to real-world environments. We train a binary classifier that distinguishes between normal and elevated stress states using resting-state EEG data recorded from 18 graduate students during a class session. The best-performing fine-tuned model achieves a balanced accuracy of 90.47% with a 5-second window, significantly outperforming traditional stress classifiers in both accuracy and inference efficiency. We further evaluate the robustness of the fine-tuned LEM under random data shuffling and reduced channel counts. These results demonstrate the capability of LEMs to effectively process real-world EEG data and highlight their potential to revolutionize brain-computer interface applications by shifting the focus from model-centric to data-centric design.
Authors: Jihai Zhang, Tianle Li, Linjie Li, Zhengyuan Yang, Yu Cheng
Abstract: Recent advancements in unified vision-language models (VLMs), which integrate both visual understanding and generation capabilities, have attracted significant attention. The underlying hypothesis is that a unified architecture with mixed training on both understanding and generation tasks can enable mutual enhancement between understanding and generation. However, this hypothesis remains underexplored in prior works on unified VLMs. To address this gap, this paper systematically investigates the generalization across understanding and generation tasks in unified VLMs. Specifically, we design a dataset closely aligned with real-world scenarios to facilitate extensive experiments and quantitative evaluations. We evaluate multiple unified VLM architectures to validate our findings. Our key findings are as follows. First, unified VLMs trained with mixed data exhibit mutual benefits in understanding and generation tasks across various architectures, and this mutual benefits can scale up with increased data. Second, better alignment between multimodal input and output spaces will lead to better generalization. Third, the knowledge acquired during generation tasks can transfer to understanding tasks, and this cross-task generalization occurs within the base language model, beyond modality adapters. Our findings underscore the critical necessity of unifying understanding and generation in VLMs, offering valuable insights for the design and optimization of unified VLMs.
Authors: Chuanhao Li, Wenbo Ye, Zhen Li, Yuwei Wu, Yunde Jia
Abstract: Compositional generalization is the ability of generalizing novel compositions from seen primitives, and has received much attention in vision-and-language (V\&L) recently. Due to the multi-modal nature of V\&L tasks, the primitives composing compositions source from different modalities, resulting in multi-sourced novel compositions. However, the generalization ability over multi-sourced novel compositions, \textit{i.e.}, multi-sourced compositional generalization (MSCG) remains unexplored. In this paper, we explore MSCG in the context of visual question answering (VQA), and propose a retrieval-augmented training framework to enhance the MSCG ability of VQA models by learning unified representations for primitives from different modalities. Specifically, semantically equivalent primitives are retrieved for each primitive in the training samples, and the retrieved features are aggregated with the original primitive to refine the model. This process helps the model learn consistent representations for the same semantic primitives across different modalities. To evaluate the MSCG ability of VQA models, we construct a new GQA-MSCG dataset based on the GQA dataset, in which samples include three types of novel compositions composed of primitives from different modalities. Experimental results demonstrate the effectiveness of the proposed framework. We release GQA-MSCG at https://github.com/NeverMoreLCH/MSCG.
Authors: Wei-Hsiang Huang, Chen-Wei Ke, Wei-Ning Chiu, Yu-Xuan Su, Chun-Chun Yang, Chieh-Yuan Cheng, Yun-Nung Chen, Pu-Jen Cheng
Abstract: Large language models (LLMs) have introduced new paradigms for recommender systems by enabling richer semantic understanding and incorporating implicit world knowledge. In this study, we propose a systematic taxonomy that classifies existing approaches into two categories: (1) Pure LLM Recommenders, which rely solely on LLMs, and (2) Augmented LLM Recommenders, which integrate additional non-LLM techniques to enhance performance. This taxonomy provides a novel lens through which to examine the evolving landscape of LLM-based recommendation. To support fair comparison, we introduce a unified evaluation platform that benchmarks representative models under consistent experimental settings, highlighting key design choices that impact effectiveness. We conclude by discussing open challenges and outlining promising directions for future research. This work offers both a comprehensive overview and practical guidance for advancing next-generation LLM-powered recommender.
Authors: Dohyeon Lee, Yeonseok Jeong, Seung-won Hwang
Abstract: Chain-of-Thought (CoT) prompting enables complex reasoning in large language models (LLMs), including applications in information retrieval (IR). However, it often leads to overthinking, where models produce excessively long and semantically redundant traces with little or no benefit. We identify two key challenges in IR: redundant trajectories that revisit similar states and misguided reasoning that diverges from user intent. To address these, we propose State Machine Reasoning (SMR), a transition-based reasoning framework composed of discrete actions (Refine, Rerank, Stop) that support early stopping and fine-grained control. Experiments on the BEIR and BRIGHT benchmarks show that SMR improves retrieval performance (nDCG@10) by 3.4% while reducing token usage by 74.4%. It generalizes across LLMs and retrievers without requiring task-specific tuning, offering a practical alternative to conventional CoT reasoning. The code and details are available at https://github.com/ldilab/SMR.
Authors: Lingkai Kong, Haichuan Wang, Tonghan Wang, Guojun Xiong, Milind Tambe
Abstract: Incorporating pre-collected offline data from a source environment can significantly improve the sample efficiency of reinforcement learning (RL), but this benefit is often challenged by discrepancies between the transition dynamics of the source and target environments. Existing methods typically address this issue by penalizing or filtering out source transitions in high dynamics-gap regions. However, their estimation of the dynamics gap often relies on KL divergence or mutual information, which can be ill-defined when the source and target dynamics have disjoint support. To overcome these limitations, we propose CompFlow, a method grounded in the theoretical connection between flow matching and optimal transport. Specifically, we model the target dynamics as a conditional flow built upon the output distribution of the source-domain flow, rather than learning it directly from a Gaussian prior. This composite structure offers two key advantages: (1) improved generalization for learning target dynamics, and (2) a principled estimation of the dynamics gap via the Wasserstein distance between source and target transitions. Leveraging our principled estimation of the dynamics gap, we further introduce an optimistic active data collection strategy that prioritizes exploration in regions of high dynamics gap, and theoretically prove that it reduces the performance disparity with the optimal policy. Empirically, CompFlow outperforms strong baselines across several RL benchmarks with shifted dynamics.
Authors: Shuyin Xia, Xiaojiang Tian, Suzhen Yuan, Jeremiah D. Deng
Abstract: High time complexity is one of the biggest challenges faced by $k$-Nearest Neighbors ($k$NN). Although current classical and quantum $k$NN algorithms have made some improvements, they still have a speed bottleneck when facing large amounts of data. To address this issue, we propose an innovative algorithm called Granular-Ball based Quantum $k$NN(GB-Q$k$NN). This approach achieves higher efficiency by first employing granular-balls, which reduces the data size needed to processed. The search process is then accelerated by adopting a Hierarchical Navigable Small World (HNSW) method. Moreover, we optimize the time-consuming steps, such as distance calculation, of the HNSW via quantization, further reducing the time complexity of the construct and search process. By combining the use of granular-balls and quantization of the HNSW method, our approach manages to take advantage of these treatments and significantly reduces the time complexity of the $k$NN-like algorithms, as revealed by a comprehensive complexity analysis.
Authors: Yuu Jinnai
Abstract: Document-level text generation tasks are known to be more difficult than sentence-level text generation tasks as they require the understanding of longer context to generate high-quality texts. In this paper, we investigate the adaption of Minimum Bayes Risk (MBR) decoding for document-level text generation tasks. MBR decoding makes use of a utility function to estimate the output with the highest expected utility from a set of candidate outputs. Although MBR decoding is shown to be effective in a wide range of sentence-level text generation tasks, its performance on document-level text generation tasks is limited as many of the utility functions are designed for evaluating the utility of sentences. To this end, we propose MBR-OT, a variant of MBR decoding using Wasserstein distance to compute the utility of a document using a sentence-level utility function. The experimental result shows that the performance of MBR-OT outperforms that of the standard MBR in document-level machine translation, text simplification, and dense image captioning tasks. Our code is available at https://github.com/jinnaiyuu/mbr-optimal-transport
Authors: Gwanghyun Kim, Xueting Li, Ye Yuan, Koki Nagano, Tianye Li, Jan Kautz, Se Young Chun, Umar Iqbal
Abstract: Estimating accurate and temporally consistent 3D human geometry from videos is a challenging problem in computer vision. Existing methods, primarily optimized for single images, often suffer from temporal inconsistencies and fail to capture fine-grained dynamic details. To address these limitations, we present GeoMan, a novel architecture designed to produce accurate and temporally consistent depth and normal estimations from monocular human videos. GeoMan addresses two key challenges: the scarcity of high-quality 4D training data and the need for metric depth estimation to accurately model human size. To overcome the first challenge, GeoMan employs an image-based model to estimate depth and normals for the first frame of a video, which then conditions a video diffusion model, reframing video geometry estimation task as an image-to-video generation problem. This design offloads the heavy lifting of geometric estimation to the image model and simplifies the video model's role to focus on intricate details while using priors learned from large-scale video datasets. Consequently, GeoMan improves temporal consistency and generalizability while requiring minimal 4D training data. To address the challenge of accurate human size estimation, we introduce a root-relative depth representation that retains critical human-scale details and is easier to be estimated from monocular inputs, overcoming the limitations of traditional affine-invariant and metric depth representations. GeoMan achieves state-of-the-art performance in both qualitative and quantitative evaluations, demonstrating its effectiveness in overcoming longstanding challenges in 3D human geometry estimation from videos.
Authors: Junyi An, Xinyu Lu, Chao Qu, Yunfei Shi, Peijia Lin, Qianwei Tang, Licheng Xu, Fenglei Cao, Yuan Qi
Abstract: SE(3)-equivariant Graph Neural Networks (GNNs) have significantly advanced molecular system modeling by employing group representations. However, their message passing processes, which rely on tensor product-based convolutions, are limited by insufficient non-linearity and incomplete group representations, thereby restricting expressiveness. To overcome these limitations, we introduce the Equivariant Spherical Transformer (EST), a novel framework that leverages a Transformer structure within the spatial domain of group representations after Fourier transform. We theoretically and empirically demonstrate that EST can encompass the function space of tensor products while achieving superior expressiveness. Furthermore, EST's equivariant inductive bias is guaranteed through a uniform sampling strategy for the Fourier transform. Our experiments demonstrate state-of-the-art performance by EST on various molecular benchmarks, including OC20 and QM9.
Authors: Pushapdeep Singh, Jyoti Nigam, Medicherla Vamsi Krishna, Arnav Bhavsar, Aditya Nigam
Abstract: While electroencephalography (EEG) has been a popular modality for neural decoding, it often involves task specific acquisition of the EEG data. This poses challenges for the development of a unified pipeline to learn embeddings for various EEG signal classification, which is often involved in various decoding tasks. Traditionally, EEG classification involves the step of signal preprocessing and the use of deep learning techniques, which are highly dependent on the number of EEG channels in each sample. However, the same pipeline cannot be applied even if the EEG data is collected for the same experiment but with different acquisition devices. This necessitates the development of a framework for learning EEG embeddings, which could be highly beneficial for tasks involving multiple EEG samples for the same task but with varying numbers of EEG channels. In this work, we propose EEG Adapter (EAD), a flexible framework compatible with any signal acquisition device. More specifically, we leverage a recent EEG foundational model with significant adaptations to learn robust representations from the EEG data for the classification task. We evaluate EAD on two publicly available datasets achieving state-of-the-art accuracies 99.33% and 92.31% on EEG-ImageNet and BrainLat respectively. This illustrates the effectiveness of the proposed framework across diverse EEG datasets containing two different perception tasks: stimulus and resting-state EEG signals. We also perform zero-shot EEG classification on EEG-ImageNet task to demonstrate the generalization capability of the proposed approach.
Authors: Pengfei Zhou, Yunlong Liu, Junli Liang, Qi Song, Xiangyang Li
Abstract: Time series forecasting with exogenous variables is a critical emerging paradigm that presents unique challenges in modeling dependencies between variables. Traditional models often struggle to differentiate between endogenous and exogenous variables, leading to inefficiencies and overfitting. In this paper, we introduce CrossLinear, a novel Linear-based forecasting model that addresses these challenges by incorporating a plug-and-play cross-correlation embedding module. This lightweight module captures the dependencies between variables with minimal computational cost and seamlessly integrates into existing neural networks. Specifically, it captures time-invariant and direct variable dependencies while disregarding time-varying or indirect dependencies, thereby mitigating the risk of overfitting in dependency modeling and contributing to consistent performance improvements. Furthermore, CrossLinear employs patch-wise processing and a global linear head to effectively capture both short-term and long-term temporal dependencies, further improving its forecasting precision. Extensive experiments on 12 real-world datasets demonstrate that CrossLinear achieves superior performance in both short-term and long-term forecasting tasks. The ablation study underscores the effectiveness of the cross-correlation embedding module. Additionally, the generalizability of this module makes it a valuable plug-in for various forecasting tasks across different domains. Codes are available at https://github.com/mumiao2000/CrossLinear.
Authors: Yuatyong Chaichana, Thanapat Trachu, Peerat Limkonchotiwat, Konpat Preechakul, Tirasan Khandhawit, Ekapol Chuangsuwanich
Abstract: In the era of large-scale training, model merging has evolved into a tool for creating multitasking models efficiently. It enables the knowledge of models to be fused, without the need for heavy computation as required in traditional multitask learning. Existing merging methods often assume that entries at identical positions in weight matrices serve the same function, enabling straightforward entry-wise comparison and merging. However, this assumption overlooks the complexity of finetuned neural networks, where neurons may develop distinct feature compositions, making direct entry-wise merging problematic. We present Decom-Renorm-Merge (DRM), a simple yet effective approach that leverages Singular Value Decomposition to decompose and coordinate weight matrices into an aligned joint space, where entry-wise merging becomes possible. We showcase the effectiveness of DRM across various settings ranging from smaller encoder-based such as ViT and DeBERTa, encoder-decoder-based such as T5, and larger decoder-based such as Llama3.1-8B. Our experimental results show that DRM outperforms several state-of-the-art merging techniques across full finetuning and low-rank adaptation settings. Moreover, our analysis reveals renormalization as the crucial component for creating a robust and even joint space for merging, significantly contributing to the method's performance.
Authors: Linjie Mu, Zhongzhen Huang, Yakun Zhu, Xiangyu Zhao, Shaoting Zhang, Xiaofan Zhang
Abstract: Effective clinical decision-making depends on iterative, multimodal reasoning across diverse sources of evidence. The recent emergence of multimodal reasoning models has significantly transformed the landscape of solving complex tasks. Although such models have achieved notable success in mathematics and science, their application to medical domains remains underexplored. In this work, we propose \textit{MedE$^2$}, a two-stage post-training pipeline that elicits and then enhances multimodal reasoning for medical domains. In Stage-I, we fine-tune models using 2,000 text-only data samples containing precisely orchestrated reasoning demonstrations to elicit reasoning behaviors. In Stage-II, we further enhance the model's reasoning capabilities using 1,500 rigorously curated multimodal medical cases, aligning model reasoning outputs with our proposed multimodal medical reasoning preference. Extensive experiments demonstrate the efficacy and reliability of \textit{MedE$^2$} in improving the reasoning performance of medical multimodal models. Notably, models trained with \textit{MedE$^2$} consistently outperform baselines across multiple medical multimodal benchmarks. Additional validation on larger models and under inference-time scaling further confirms the robustness and practical utility of our approach.
Authors: Yiming Lei, Zhizheng Yang, Zeming Liu, Haitao Leng, Shaoguo Liu, Tingting Gao, Qingjie Liu, Yunhong Wang
Abstract: Multi-modal large language models have demonstrated remarkable zero-shot abilities and powerful image-understanding capabilities. However, the existing open-source multi-modal models suffer from the weak capability of multi-turn interaction, especially for long contexts. To address the issue, we first introduce a context modeling module, termed ContextQFormer, which utilizes a memory block to enhance the presentation of contextual information. Furthermore, to facilitate further research, we carefully build a new multi-turn multi-modal dialogue dataset (TMDialog) for pre-training, instruction-tuning, and evaluation, which will be open-sourced lately. Compared with other multi-modal dialogue datasets, TMDialog contains longer conversations, which supports the research of multi-turn multi-modal dialogue. In addition, ContextQFormer is compared with three baselines on TMDialog and experimental results illustrate that ContextQFormer achieves an improvement of 2%-4% in available rate over baselines.
Authors: Seung Gyu Jeong, Seong Eun Kim
Abstract: Auscultation is crucial for diagnosing lung diseases. The COVID-19 pandemic has revealed the limitations of traditional, in-person lung sound assessments. To overcome these issues, advancements in digital stethoscopes and artificial intelligence (AI) have led to the development of new diagnostic methods. In this context, our study aims to use smartphone microphones to record and analyze lung sounds. We faced two major challenges: the difference in audio style between electronic stethoscopes and smartphone microphones, and the variability among patients. To address these challenges, we developed a method called Patient Domain Supervised Contrastive Learning (PD-SCL). By integrating this method with the Audio Spectrogram Transformer (AST) model, we significantly improved its performance by 2.4\% compared to the original AST model. This progress demonstrates that smartphones can effectively diagnose lung sounds, addressing inconsistencies in patient data and showing potential for broad use beyond traditional clinical settings. Our research contributes to making lung disease detection more accessible in the post-COVID-19 world.
Authors: Tongtong Su, Chengyu Wang, Jun Huang, Dongming Lu
Abstract: Appearance editing according to user needs is a pivotal task in video editing. Existing text-guided methods often lead to ambiguities regarding user intentions and restrict fine-grained control over editing specific aspects of objects. To overcome these limitations, this paper introduces a novel approach named {Zero-to-Hero}, which focuses on reference-based video editing that disentangles the editing process into two distinct problems. It achieves this by first editing an anchor frame to satisfy user requirements as a reference image and then consistently propagating its appearance across other frames. We leverage correspondence within the original frames to guide the attention mechanism, which is more robust than previously proposed optical flow or temporal modules in memory-friendly video generative models, especially when dealing with objects exhibiting large motions. It offers a solid ZERO-shot initialization that ensures both accuracy and temporal consistency. However, intervention in the attention mechanism results in compounded imaging degradation with over-saturated colors and unknown blurring issues. Starting from Zero-Stage, our Hero-Stage Holistically learns a conditional generative model for vidEo RestOration. To accurately evaluate the consistency of the appearance, we construct a set of videos with multiple appearances using Blender, enabling a fine-grained and deterministic evaluation. Our method outperforms the best-performing baseline with a PSNR improvement of 2.6 dB. The project page is at https://github.com/Tonniia/Zero2Hero.
Authors: Zhe Ye, Zhengxu Yan, Jingxuan He, Timothe Kasriel, Kaiyu Yang, Dawn Song
Abstract: Large language models (LLMs) are increasingly integrated in software development, but ensuring correctness in LLM-generated code remains challenging and often requires costly manual review. Verifiable code generation -- jointly generating code, specifications, and proofs of code-specification alignment -- offers a promising path to address this limitation and further unleash LLMs' benefits in coding. Yet, there exists a significant gap in evaluation: current benchmarks often lack support for end-to-end verifiable code generation. In this paper, we introduce Verina (Verifiable Code Generation Arena), a high-quality benchmark enabling a comprehensive and modular evaluation of code, specification, and proof generation as well as their compositions. Verina consists of 189 manually curated coding tasks in Lean, with detailed problem descriptions, reference implementations, formal specifications, and extensive test suites. Our extensive evaluation of state-of-the-art LLMs reveals significant challenges in verifiable code generation, especially in proof generation, underscoring the need for improving LLM-based theorem provers in verification domains. The best model, OpenAI o4-mini, generates only 61.4% correct code, 51.0% sound and complete specifications, and 3.6% successful proofs, with one trial per task. We hope Verina will catalyze progress in verifiable code generation by providing a rigorous and comprehensive benchmark. We release our dataset on https://huggingface.co/datasets/sunblaze-ucb/verina and our evaluation code on https://github.com/sunblaze-ucb/verina.
URLs: https://huggingface.co/datasets/sunblaze-ucb/verina, https://github.com/sunblaze-ucb/verina.
Authors: Jeongsol Kim, Yeobin Hong, Jong Chul Ye
Abstract: Recent inversion-free, flow-based image editing methods such as FlowEdit leverages a pre-trained noise-to-image flow model such as Stable Diffusion 3, enabling text-driven manipulation by solving an ordinary differential equation (ODE). While the lack of exact latent inversion is a core advantage of these methods, it often results in unstable editing trajectories and poor source consistency. To address this limitation, we propose FlowAlign, a novel inversion-free flow-based framework for consistent image editing with principled trajectory control. FlowAlign introduces a flow-matching loss as a regularization mechanism to promote smoother and more stable trajectories during the editing process. Notably, the flow-matching loss is shown to explicitly balance semantic alignment with the edit prompt and structural consistency with the source image along the trajectory. Furthermore, FlowAlign naturally supports reverse editing by simply reversing the ODE trajectory, highlighting the reversible and consistent nature of the transformation. Extensive experiments demonstrate that FlowAlign outperforms existing methods in both source preservation and editing controllability.
Authors: Antonio D'Orazio, Maria Rosaria Briglia, Donato Crisostomi, Dario Loi, Emanuele Rodol\`a, Iacopo Masi
Abstract: CLIP is a discriminative model trained to align images and text in a shared embedding space. Due to its multimodal structure, it serves as the backbone of many generative pipelines, where a decoder is trained to map from the shared space back to images. In this work, we show that image synthesis is nevertheless possible using CLIP alone -- without any decoder, training, or fine-tuning. Our approach optimizes a frequency-aware implicit neural representation that encourages coarse-to-fine generation by stratifying frequencies across network layers. To stabilize this inverse mapping, we introduce adversarially robust initialization, a lightweight Orthogonal Procrustes projection to align local text and image embeddings, and a blending loss that anchors outputs to natural image statistics. Without altering CLIP's weights, this framework unlocks capabilities such as text-to-image generation, style transfer, and image reconstruction. These findings suggest that discriminative models may hold untapped generative potential, hidden in plain sight.
Authors: Le Yang, Vincent Y. F. Tan, Wang Chi Cheung
Abstract: We study the best arm identification (BAI) problem with potentially biased offline data in the fixed confidence setting, which commonly arises in real-world scenarios such as clinical trials. We prove an impossibility result for adaptive algorithms without prior knowledge of the bias bound between online and offline distributions. To address this, we propose the LUCB-H algorithm, which introduces adaptive confidence bounds by incorporating an auxiliary bias correction to balance offline and online data within the LUCB framework. Theoretical analysis shows that LUCB-H matches the sample complexity of standard LUCB when offline data is misleading and significantly outperforms it when offline data is helpful. We also derive an instance-dependent lower bound that matches the upper bound of LUCB-H in certain scenarios. Numerical experiments further demonstrate the robustness and adaptability of LUCB-H in effectively incorporating offline data.
Authors: Tian Tian, Chunyan Miao, Hangwei Qian
Abstract: Contrastive learning has emerged as a competent approach for unsupervised representation learning. However, the design of an optimal augmentation strategy, although crucial for contrastive learning, is less explored for time series classification tasks. Existing predefined time-domain augmentation methods are primarily adopted from vision and are not specific to time series data. Consequently, this cross-modality incompatibility may distort the semantically relevant information of time series by introducing mismatched patterns into the data. To address this limitation, we present a novel perspective from the frequency domain and identify three advantages for downstream classification: global, independent, and compact. To fully utilize the three properties, we propose the lightweight yet effective Frequency Refined Augmentation (FreRA) tailored for time series contrastive learning on classification tasks, which can be seamlessly integrated with contrastive learning frameworks in a plug-and-play manner. Specifically, FreRA automatically separates critical and unimportant frequency components. Accordingly, we propose semantic-aware Identity Modification and semantic-agnostic Self-adaptive Modification to protect semantically relevant information in the critical frequency components and infuse variance into the unimportant ones respectively. Theoretically, we prove that FreRA generates semantic-preserving views. Empirically, we conduct extensive experiments on two benchmark datasets, including UCR and UEA archives, as well as five large-scale datasets on diverse applications. FreRA consistently outperforms ten leading baselines on time series classification, anomaly detection, and transfer learning tasks, demonstrating superior capabilities in contrastive representation learning and generalization in transfer learning scenarios across diverse datasets.
Authors: Gabriele Sarti, Vil\'em Zouhar, Malvina Nissim, Arianna Bisazza
Abstract: Word-level quality estimation (WQE) aims to automatically identify fine-grained error spans in machine-translated outputs and has found many uses, including assisting translators during post-editing. Modern WQE techniques are often expensive, involving prompting of large language models or ad-hoc training on large amounts of human-labeled data. In this work, we investigate efficient alternatives exploiting recent advances in language model interpretability and uncertainty quantification to identify translation errors from the inner workings of translation models. In our evaluation spanning 14 metrics across 12 translation directions, we quantify the impact of human label variation on metric performance by using multiple sets of human labels. Our results highlight the untapped potential of unsupervised metrics, the shortcomings of supervised methods when faced with label uncertainty, and the brittleness of single-annotator evaluation practices.
Authors: Yilong Li, Chen Qian, Yu Xia, Ruijie Shi, Yufan Dang, Zihao Xie, Ziming You, Weize Chen, Cheng Yang, Weichuan Liu, Ye Tian, Xuantang Xiong, Lei Han, Zhiyuan Liu, Maosong Sun
Abstract: Large Language Model-based multi-agent systems (MAS) have shown remarkable progress in solving complex tasks through collaborative reasoning and inter-agent critique. However, existing approaches typically treat each task in isolation, resulting in redundant computations and limited generalization across structurally similar tasks. To address this, we introduce multi-agent cross-task experiential learning (MAEL), a novel framework that endows LLM-driven agents with explicit cross-task learning and experience accumulation. We model the task-solving workflow on a graph-structured multi-agent collaboration network, where agents propagate information and coordinate via explicit connectivity. During the experiential learning phase, we quantify the quality for each step in the task-solving workflow and store the resulting rewards along with the corresponding inputs and outputs into each agent's individual experience pool. During inference, agents retrieve high-reward, task-relevant experiences as few-shot examples to enhance the effectiveness of each reasoning step, thereby enabling more accurate and efficient multi-agent collaboration. Experimental results on diverse datasets demonstrate that MAEL empowers agents to learn from prior task experiences effectively-achieving faster convergence and producing higher-quality solutions on current tasks.
Authors: Jinglong Gao, Xiao Ding, Lingxiao Zou, Bibo Cai, Bing Qin, Ting Liu
Abstract: Recent studies provide large language models (LLMs) with textual task-solving experiences via prompts to improve their performance. However, previous methods rely on substantial human labor or time to gather such experiences for each task, which is impractical given the growing variety of task types in user queries to LLMs. To address this issue, we design an autonomous experience transfer framework to explore whether LLMs can mimic human cognitive intelligence to autonomously transfer experience from existing source tasks to newly encountered target tasks. This not only allows the acquisition of experience without extensive costs of previous methods, but also offers a novel path for the generalization of LLMs. Experimental results on 13 datasets demonstrate that our framework effectively improves the performance of LLMs. Furthermore, we provide a detailed analysis of each module in the framework.
Authors: Run Hao, Peng Ying
Abstract: The rise of text-to-image (T2I) models has enabled the synthesis of photorealistic human portraits, raising serious concerns about identity misuse and the robustness of AIGC detectors. In this work, we propose an automated adversarial prompt generation framework that leverages a grammar tree structure and a variant of the Monte Carlo tree search algorithm to systematically explore the semantic prompt space. Our method generates diverse, controllable prompts that consistently evade both open-source and commercial AIGC detectors. Extensive experiments across multiple T2I models validate its effectiveness, and the approach ranked first in a real-world adversarial AIGC detection competition. Beyond attack scenarios, our method can also be used to construct high-quality adversarial datasets, providing valuable resources for training and evaluating more robust AIGC detection and defense systems.
Authors: Lifan Zhao, Yanyan Shen, Zhaoyang Liu, Xue Wang, Jiaji Deng
Abstract: Scaling laws motivate the development of Time Series Foundation Models (TSFMs) that pre-train vast parameters and achieve remarkable zero-shot forecasting performance. Surprisingly, even after fine-tuning, TSFMs cannot consistently outperform smaller, specialized models trained on full-shot downstream data. A key question is how to realize effective adaptation of TSFMs for a target forecasting task. Through empirical studies on various TSFMs, the pre-trained models often exhibit inherent sparsity and redundancy in computation, suggesting that TSFMs have learned to activate task-relevant network substructures to accommodate diverse forecasting tasks. To preserve this valuable prior knowledge, we propose a structured pruning method to regularize the subsequent fine-tuning process by focusing it on a more relevant and compact parameter space. Extensive experiments on seven TSFMs and six benchmarks demonstrate that fine-tuning a smaller, pruned TSFM significantly improves forecasting performance compared to fine-tuning original models. This "prune-then-finetune" paradigm often enables TSFMs to achieve state-of-the-art performance and surpass strong specialized baselines.
Authors: Wenhao Xu, Shuchen Zheng, Changwei Wang, Zherui Zhang, Chuan Ren, Rongtao Xu, Shibiao Xu
Abstract: Infrared small target detection (ISTD) is vital for long-range surveillance in military, maritime, and early warning applications. ISTD is challenged by targets occupying less than 0.15% of the image and low distinguishability from complex backgrounds. Existing deep learning methods often suffer from information loss during downsampling and inefficient global context modeling. This paper presents SAMamba, a novel framework integrating SAM2's hierarchical feature learning with Mamba's selective sequence modeling. Key innovations include: (1) A Feature Selection Adapter (FS-Adapter) for efficient natural-to-infrared domain adaptation via dual-stage selection (token-level with a learnable task embedding and channel-wise adaptive transformations); (2) A Cross-Channel State-Space Interaction (CSI) module for efficient global context modeling with linear complexity using selective state space modeling; and (3) A Detail-Preserving Contextual Fusion (DPCF) module that adaptively combines multi-scale features with a gating mechanism to balance high-resolution and low-resolution feature contributions. SAMamba addresses core ISTD challenges by bridging the domain gap, maintaining fine-grained details, and efficiently modeling long-range dependencies. Experiments on NUAA-SIRST, IRSTD-1k, and NUDT-SIRST datasets show SAMamba significantly outperforms state-of-the-art methods, especially in challenging scenarios with heterogeneous backgrounds and varying target scales. Code: https://github.com/zhengshuchen/SAMamba.
Authors: Hao Lu, Yanchi Gu, Haoyuan Huang, Yulin Zhou, Ningxin Zhu, Chen Li
Abstract: The integration of Monte Carlo Tree Search (MCTS) with Large Language Models (LLMs) has demonstrated significant success in structured, problem-oriented tasks. However, applying these methods to open-ended dialogues, such as those in psychological counseling, presents unique challenges. Unlike tasks with objective correctness, success in therapeutic conversations depends on subjective factors like empathetic engagement, ethical adherence, and alignment with human preferences, for which strict "correctness" criteria are ill-defined. Existing result-oriented MCTS approaches can therefore produce misaligned responses. To address this, we introduce MCTSr-Zero, an MCTS framework designed for open-ended, human-centric dialogues. Its core innovation is "domain alignment", which shifts the MCTS search objective from predefined end-states towards conversational trajectories that conform to target domain principles (e.g., empathy in counseling). Furthermore, MCTSr-Zero incorporates "Regeneration" and "Meta-Prompt Adaptation" mechanisms to substantially broaden exploration by allowing the MCTS to consider fundamentally different initial dialogue strategies. We evaluate MCTSr-Zero in psychological counseling by generating multi-turn dialogue data, which is used to fine-tune an LLM, PsyLLM. We also introduce PsyEval, a benchmark for assessing multi-turn psychological counseling dialogues. Experiments demonstrate that PsyLLM achieves state-of-the-art performance on PsyEval and other relevant metrics, validating MCTSr-Zero's effectiveness in generating high-quality, principle-aligned conversational data for human-centric domains and addressing the LLM challenge of consistently adhering to complex psychological standards.
Authors: Lingkai Meng, Yu Shao, Long Yuan, Longbin Lai, Peng Cheng, Wenyuan Yu, Wenjie Zhang, Xuemin Lin, Jingren Zhou
Abstract: Usability evaluation is critical to the impact and adoption of open source software (OSS), yet traditional methods relying on human evaluators suffer from high costs and limited scalability. To address these limitations, we introduce OSS-UAgent, an automated, configurable, and interactive agent-based usability evaluation framework specifically designed for open source software. Our framework employs intelligent agents powered by large language models (LLMs) to simulate developers performing programming tasks across various experience levels (from Junior to Expert). By dynamically constructing platform-specific knowledge bases, OSS-UAgent ensures accurate and context-aware code generation. The generated code is automatically evaluated across multiple dimensions, including compliance, correctness, and readability, providing a comprehensive measure of the software's usability. Additionally, our demonstration showcases OSS-UAgent's practical application in evaluating graph analytics platforms, highlighting its effectiveness in automating usability evaluation.
Authors: Zonglin Yang, Zhexuan Gu, Houduo Qi, Yancheng Yuan
Abstract: Reinforcement learning from human feedback (RLHF) is an essential technique for ensuring that large language models (LLMs) are aligned with human values and preferences during the post-training phase. As an effective RLHF approach, group relative policy optimization (GRPO) has demonstrated success in many LLM-based applications. However, efficient GRPO-based RLHF training remains a challenge. Recent studies reveal that a higher reward variance of the initial policy model leads to faster RLHF training. Inspired by this finding, we propose a practical reward adjustment model to accelerate RLHF training by provably increasing the reward variance and preserving the relative preferences and reward expectation. Our reward adjustment method inherently poses a nonconvex optimization problem, which is NP-hard to solve in general. To overcome the computational challenges, we design a novel $O(n \log n)$ algorithm to find a global solution of the nonconvex reward adjustment model by explicitly characterizing the extreme points of the feasible set. As an important application, we naturally integrate this reward adjustment model into the GRPO algorithm, leading to a more efficient GRPO with reward variance increase (GRPOVI) algorithm for RLHF training. As an interesting byproduct, we provide an indirect explanation for the empirical effectiveness of GRPO with rule-based reward for RLHF training, as demonstrated in DeepSeek-R1. Experiment results demonstrate that the GRPOVI algorithm can significantly improve the RLHF training efficiency compared to the original GRPO algorithm.
Authors: Pascal J. Sager, Ashwini Kamaraj, Benjamin F. Grewe, Thilo Stadelmann
Abstract: We present the methodology and results of the Deep Retrieval team for subtask 4b of the CLEF CheckThat! 2025 competition, which focuses on retrieving relevant scientific literature for given social media posts. To address this task, we propose a hybrid retrieval pipeline that combines lexical precision, semantic generalization, and deep contextual re-ranking, enabling robust retrieval that bridges the informal-to-formal language gap. Specifically, we combine BM25-based keyword matching with a FAISS vector store using a fine-tuned INF-Retriever-v1 model for dense semantic retrieval. BM25 returns the top 30 candidates, and semantic search yields 100 candidates, which are then merged and re-ranked via a large language model (LLM)-based cross-encoder. Our approach achieves a mean reciprocal rank at 5 (MRR@5) of 76.46% on the development set and 66.43% on the hidden test set, securing the 1st position on the development leaderboard and ranking 3rd on the test leaderboard (out of 31 teams), with a relative performance gap of only 2 percentage points compared to the top-ranked system. We achieve this strong performance by running open-source models locally and without external training data, highlighting the effectiveness of a carefully designed and fine-tuned retrieval pipeline.
Authors: Chunlong Xie, Jialing He, Shangwei Guo, Jiacheng Wang, Shudong Zhang, Tianwei Zhang, Tao Xiang
Abstract: We present Adversarial Object Fusion (AdvOF), a novel attack framework targeting vision-and-language navigation (VLN) agents in service-oriented environments by generating adversarial 3D objects. While foundational models like Large Language Models (LLMs) and Vision Language Models (VLMs) have enhanced service-oriented navigation systems through improved perception and decision-making, their integration introduces vulnerabilities in mission-critical service workflows. Existing adversarial attacks fail to address service computing contexts, where reliability and quality-of-service (QoS) are paramount. We utilize AdvOF to investigate and explore the impact of adversarial environments on the VLM-based perception module of VLN agents. In particular, AdvOF first precisely aggregates and aligns the victim object positions in both 2D and 3D space, defining and rendering adversarial objects. Then, we collaboratively optimize the adversarial object with regularization between the adversarial and victim object across physical properties and VLM perceptions. Through assigning importance weights to varying views, the optimization is processed stably and multi-viewedly by iterative fusions from local updates and justifications. Our extensive evaluations demonstrate AdvOF can effectively degrade agent performance under adversarial conditions while maintaining minimal interference with normal navigation tasks. This work advances the understanding of service security in VLM-powered navigation systems, providing computational foundations for robust service composition in physical-world deployments.
Authors: Jianlin Ye, Savvas Papaioannou, Panayiotis Kolios
Abstract: Path planning is a fundamental capability of autonomous Unmanned Aerial Vehicles (UAVs), enabling them to efficiently navigate toward a target region or explore complex environments while avoiding obstacles. Traditional pathplanning methods, such as Rapidly-exploring Random Trees (RRT), have proven effective but often encounter significant challenges. These include high search space complexity, suboptimal path quality, and slow convergence, issues that are particularly problematic in high-stakes applications like disaster response, where rapid and efficient planning is critical. To address these limitations and enhance path-planning efficiency, we propose Vision Language Model RRT (VLM-RRT), a hybrid approach that integrates the pattern recognition capabilities of Vision Language Models (VLMs) with the path-planning strengths of RRT. By leveraging VLMs to provide initial directional guidance based on environmental snapshots, our method biases sampling toward regions more likely to contain feasible paths, significantly improving sampling efficiency and path quality. Extensive quantitative and qualitative experiments with various state-of-the-art VLMs demonstrate the effectiveness of this proposed approach.
Authors: Spyros Barbakos, Charalampos Antoniadis, Gerasimos Potamianos, Gianluca Setti
Abstract: Video consumption is a key part of daily life, but watching entire videos can be tedious. To address this, researchers have explored video summarization and highlight detection to identify key video segments. While some works combine video frames and transcripts, and others tackle video summarization and highlight detection using Reinforcement Learning (RL), no existing work, to the best of our knowledge, integrates both modalities within an RL framework. In this paper, we propose a multimodal pipeline that leverages video frames and their corresponding transcripts to generate a more condensed version of the video and detect highlights using a modality fusion mechanism. The pipeline is trained within an RL framework, which rewards the model for generating diverse and representative summaries while ensuring the inclusion of video segments with meaningful transcript content. The unsupervised nature of the training allows for learning from large-scale unannotated datasets, overcoming the challenge posed by the limited size of existing annotated datasets. Our experiments show that using the transcript in video summarization and highlight detection achieves superior results compared to relying solely on the visual content of the video.
Authors: Haokun Chen, Yueqi Zhang, Yuan Bi, Yao Zhang, Tong Liu, Jinhe Bi, Jian Lan, Jindong Gu, Claudia Grosser, Denis Krompass, Nassir Navab, Volker Tresp
Abstract: In recent years, Large Language Models (LLMs) have achieved remarkable advancements, drawing significant attention from the research community. Their capabilities are largely attributed to large-scale architectures, which require extensive training on massive datasets. However, such datasets often contain sensitive or copyrighted content sourced from the public internet, raising concerns about data privacy and ownership. Regulatory frameworks, such as the General Data Protection Regulation (GDPR), grant individuals the right to request the removal of such sensitive information. This has motivated the development of machine unlearning algorithms that aim to remove specific knowledge from models without the need for costly retraining. Despite these advancements, evaluating the efficacy of unlearning algorithms remains a challenge due to the inherent complexity and generative nature of LLMs. In this work, we introduce a comprehensive auditing framework for unlearning evaluation, comprising three benchmark datasets, six unlearning algorithms, and five prompt-based auditing methods. By using various auditing algorithms, we evaluate the effectiveness and robustness of different unlearning strategies. To explore alternatives beyond prompt-based auditing, we propose a novel technique that leverages intermediate activation perturbations, addressing the limitations of auditing methods that rely solely on model inputs and outputs.
Authors: Maged S. Al-Shaibani, Moataz Ahmed
Abstract: Large Language Models (LLMs) have achieved unprecedented capabilities in generating human-like text, posing subtle yet significant challenges for information integrity across critical domains, including education, social media, and academia, enabling sophisticated misinformation campaigns, compromising healthcare guidance, and facilitating targeted propaganda. This challenge becomes severe, particularly in under-explored and low-resource languages like Arabic. This paper presents a comprehensive investigation of Arabic machine-generated text, examining multiple generation strategies (generation from the title only, content-aware generation, and text refinement) across diverse model architectures (ALLaM, Jais, Llama, and GPT-4) in academic, and social media domains. Our stylometric analysis reveals distinctive linguistic patterns differentiating human-written from machine-generated Arabic text across these varied contexts. Despite their human-like qualities, we demonstrate that LLMs produce detectable signatures in their Arabic outputs, with domain-specific characteristics that vary significantly between different contexts. Based on these insights, we developed BERT-based detection models that achieved exceptional performance in formal contexts (up to 99.9\% F1-score) with strong precision across model architectures. Our cross-domain analysis confirms generalization challenges previously reported in the literature. To the best of our knowledge, this work represents the most comprehensive investigation of Arabic machine-generated text to date, uniquely combining multiple prompt generation methods, diverse model architectures, and in-depth stylometric analysis across varied textual domains, establishing a foundation for developing robust, linguistically-informed detection systems essential for preserving information integrity in Arabic-language contexts.
Authors: Yong Zhang, Yanwen Huang, Ning Cheng, Yang Guo, Yun Zhu, Yanmeng Wang, Shaojun Wang, Jing Xiao
Abstract: Retrieval-augmented generation (RAG) enhances large language models (LLMs) with external context, but retrieved passages are often lengthy, noisy, or exceed input limits. Existing compression methods typically require supervised training of dedicated compression models, increasing cost and reducing portability. We propose Sentinel, a lightweight sentence-level compression framework that reframes context filtering as an attention-based understanding task. Rather than training a compression model, Sentinel probes decoder attention from an off-the-shelf 0.5B proxy LLM using a lightweight classifier to identify sentence relevance. Empirically, we find that query-context relevance estimation is consistent across model scales, with 0.5B proxies closely matching the behaviors of larger models. On the LongBench benchmark, Sentinel achieves up to 5$\times$ compression while matching the QA performance of 7B-scale compression systems. Our results suggest that probing native attention signals enables fast, effective, and question-aware context compression. Code available at: https://github.com/yzhangchuck/Sentinel.
Authors: Evangelos Charalampakis, Vasileios Mygdalis, Ioannis Pitas
Abstract: This work explores the application of Federated Learning (FL) in Unsupervised Semantic image Segmentation (USS). Recent USS methods extract pixel-level features using frozen visual foundation models and refine them through self-supervised objectives that encourage semantic grouping. These features are then grouped to semantic clusters to produce segmentation masks. Extending these ideas to federated settings requires feature representation and cluster centroid alignment across distributed clients -- an inherently difficult task under heterogeneous data distributions in the absence of supervision. To address this, we propose FUSS Federated Unsupervised image Semantic Segmentation) which is, to our knowledge, the first framework to enable fully decentralized, label-free semantic segmentation training. FUSS introduces novel federation strategies that promote global consistency in feature and prototype space, jointly optimizing local segmentation heads and shared semantic centroids. Experiments on both benchmark and real-world datasets, including binary and multi-class segmentation tasks, show that FUSS consistently outperforms local-only client trainings as well as extensions of classical FL algorithms under varying client data distributions. To support reproducibility, full code will be released upon manuscript acceptance.
Authors: James Xu Zhao, Jimmy Z. J. Liu, Bryan Hooi, See-Kiong Ng
Abstract: Large language models (LLMs) are widely used for long-form text generation. However, factual errors in the responses would undermine their reliability. Despite growing attention to LLM factuality, the effect of response length on factuality remains underexplored. In this work, we systematically investigate this relationship by first introducing an automatic and bi-level long-form factuality evaluation framework, which achieves high agreement with human annotations while being cost-effective. Using this framework, we conduct controlled experiments and find that longer responses exhibit lower factual precision, confirming the presence of length bias. To explain this phenomenon, we empirically examine three hypotheses: error propagation, long context, and facts exhaustion. Our results reveal that facts exhaustion, where the model gradually exhausts more reliable knowledge, is the primary cause of factual degradation, rather than the other two hypotheses.
Authors: Nimrod Shabtay, Zvi Kons, Avihu Dekel, Hagai Aronowitz, Ron Hoory, Assaf Arbelle
Abstract: Question answering (QA) systems are designed to answer natural language questions. Visual QA (VQA) and Spoken QA (SQA) systems extend the textual QA system to accept visual and spoken input respectively. This work aims to create a system that enables user interaction through both speech and images. That is achieved through the fusion of text, speech, and image modalities to tackle the task of spoken VQA (SVQA). The resulting multi-modal model has textual, visual, and spoken inputs and can answer spoken questions on images. Training and evaluating SVQA models requires a dataset for all three modalities, but no such dataset currently exists. We address this problem by synthesizing VQA datasets using two zero-shot TTS models. Our initial findings indicate that a model trained only with synthesized speech nearly reaches the performance of the upper-bounding model trained on textual QAs. In addition, we show that the choice of the TTS model has a minor impact on accuracy.
Authors: Yixin Ren, Chenghou Jin, Yewei Xia, Li Ke, Longtao Huang, Hui Xue, Hao Zhang, Jihong Guan, Shuigeng Zhou
Abstract: Determining conditional independence (CI) relationships between random variables is a fundamental yet challenging task in machine learning and statistics, especially in high-dimensional settings. Existing generative model-based CI testing methods, such as those utilizing generative adversarial networks (GANs), often struggle with undesirable modeling of conditional distributions and training instability, resulting in subpar performance. To address these issues, we propose a novel CI testing method via score-based generative modeling, which achieves precise Type I error control and strong testing power. Concretely, we first employ a sliced conditional score matching scheme to accurately estimate conditional score and use Langevin dynamics conditional sampling to generate null hypothesis samples, ensuring precise Type I error control. Then, we incorporate a goodness-of-fit stage into the method to verify generated samples and enhance interpretability in practice. We theoretically establish the error bound of conditional distributions modeled by score-based generative models and prove the validity of our CI tests. Extensive experiments on both synthetic and real-world datasets show that our method significantly outperforms existing state-of-the-art methods, providing a promising way to revitalize generative model-based CI testing.
Authors: Weizhe Kong, Xiao Wang, Ruichong Gao, Chenglong Li, Yu Zhang, Xing Yang, Yaowei Wang, Jin Tang
Abstract: Pedestrian Attribute Recognition (PAR) is an indispensable task in human-centered research and has made great progress in recent years with the development of deep neural networks. However, the potential vulnerability and anti-interference ability have still not been fully explored. To bridge this gap, this paper proposes the first adversarial attack and defense framework for pedestrian attribute recognition. Specifically, we exploit both global- and patch-level attacks on the pedestrian images, based on the pre-trained CLIP-based PAR framework. It first divides the input pedestrian image into non-overlapping patches and embeds them into feature embeddings using a projection layer. Meanwhile, the attribute set is expanded into sentences using prompts and embedded into attribute features using a pre-trained CLIP text encoder. A multi-modal Transformer is adopted to fuse the obtained vision and text tokens, and a feed-forward network is utilized for attribute recognition. Based on the aforementioned PAR framework, we adopt the adversarial semantic and label-perturbation to generate the adversarial noise, termed ASL-PAR. We also design a semantic offset defense strategy to suppress the influence of adversarial attacks. Extensive experiments conducted on both digital domains (i.e., PETA, PA100K, MSP60K, RAPv2) and physical domains fully validated the effectiveness of our proposed adversarial attack and defense strategies for the pedestrian attribute recognition. The source code of this paper will be released on https://github.com/Event-AHU/OpenPAR.
Authors: Abhirup Chakravarty, Mark Brenchley, Trevor Breakspear, Ian Lewin, Yan Huang
Abstract: A key ethical challenge in Automated Essay Scoring (AES) is ensuring that scores are only released when they meet high reliability standards. Confidence modelling addresses this by assigning a reliability estimate measure, in the form of a confidence score, to each automated score. In this study, we frame confidence estimation as a classification task: predicting whether an AES-generated score correctly places a candidate in the appropriate CEFR level. While this is a binary decision, we leverage the inherent granularity of the scoring domain in two ways. First, we reformulate the task as an n-ary classification problem using score binning. Second, we introduce a set of novel Kernel Weighted Ordinal Categorical Cross Entropy (KWOCCE) loss functions that incorporate the ordinal structure of CEFR labels. Our best-performing model achieves an F1 score of 0.97, and enables the system to release 47% of scores with 100% CEFR agreement and 99% with at least 95% CEFR agreement -compared to approximately 92% (approx.) CEFR agreement from the standalone AES model where we release all AM predicted scores.
Authors: Matteo Gallici, Haitz S\'aez de Oc\'ariz Borde
Abstract: Fine-tuning pre-trained generative models with Reinforcement Learning (RL) has emerged as an effective approach for aligning outputs more closely with nuanced human preferences. In this paper, we investigate the application of Group Relative Policy Optimization (GRPO) to fine-tune next-scale visual autoregressive (VAR) models. Our empirical results demonstrate that this approach enables alignment to intricate reward signals derived from aesthetic predictors and CLIP embeddings, significantly enhancing image quality and enabling precise control over the generation style. Interestingly, by leveraging CLIP, our method can help VAR models generalize beyond their initial ImageNet distribution: through RL-driven exploration, these models can generate images aligned with prompts referencing image styles that were absent during pre-training. In summary, we show that RL-based fine-tuning is both efficient and effective for VAR models, benefiting particularly from their fast inference speeds, which are advantageous for online sampling, an aspect that poses significant challenges for diffusion-based alternatives.
Authors: Chetan Verma, Aditya Srinivas Timmaraju, Cho Jui-Hsieh, Suyash Damle, Ngot Bui, Yang Zhang, Wen Chen, Xin Liu, Prateek Jain, Inderjit S Dhillon
Abstract: Industry-grade ML models are carefully designed to meet rapidly evolving serving constraints, which requires significant resources for model development. In this paper, we propose MatTA, a framework for training multiple accurate Student models using a novel Teacher-TA-Student recipe. TA models are larger versions of the Student models with higher capacity, and thus allow Student models to better relate to the Teacher model and also bring in more domain-specific expertise. Furthermore, multiple accurate Student models can be extracted from the TA model. Therefore, despite only one training run, our methodology provides multiple servable options to trade off accuracy for lower serving cost. We demonstrate the proposed method, MatTA, on proprietary datasets and models. Its practical efficacy is underscored by live A/B tests within a production ML system, demonstrating 20% improvement on a key metric. We also demonstrate our method on GPT-2 Medium, a public model, and achieve relative improvements of over 24% on SAT Math and over 10% on the LAMBADA benchmark.
Authors: Sheng Ouyang, Yulan Hu, Ge Chen, Qingyang Li, Fuzheng Zhang, Yong Liu
Abstract: Rewards serve as proxies for human preferences and play a crucial role in Reinforcement Learning from Human Feedback (RLHF). However, if these rewards are inherently imperfect, exhibiting various biases, they can adversely affect the alignment of large language models (LLMs). In this paper, we collectively define the various biases present in rewards as the problem of reward unfairness. We propose a bias-agnostic method to address the issue of reward fairness from a resource allocation perspective, without specifically designing for each type of bias, yet effectively mitigating them. Specifically, we model preference learning as a resource allocation problem, treating rewards as resources to be allocated while considering the trade-off between utility and fairness in their distribution. We propose two methods, Fairness Regularization and Fairness Coefficient, to achieve fairness in rewards. We apply our methods in both verification and reinforcement learning scenarios to obtain a fairness reward model and a policy model, respectively. Experiments conducted in these scenarios demonstrate that our approach aligns LLMs with human preferences in a more fair manner.
Authors: Xu Shen, Yixin Liu, Yiwei Dai, Yili Wang, Rui Miao, Yue Tan, Shirui Pan, Xin Wang
Abstract: The communication topology in large language model-based multi-agent systems fundamentally governs inter-agent collaboration patterns, critically shaping both the efficiency and effectiveness of collective decision-making. While recent studies for communication topology automated design tend to construct sparse structures for efficiency, they often overlook why and when sparse and dense topologies help or hinder collaboration. In this paper, we present a causal framework to analyze how agent outputs, whether correct or erroneous, propagate under topologies with varying sparsity. Our empirical studies reveal that moderately sparse topologies, which effectively suppress error propagation while preserving beneficial information diffusion, typically achieve optimal task performance. Guided by this insight, we propose a novel topology design approach, EIB-leanrner, that balances error suppression and beneficial information propagation by fusing connectivity patterns from both dense and sparse graphs. Extensive experiments show the superior effectiveness, communication cost, and robustness of EIB-leanrner.
Authors: Alexandra G. Roberts, Ha M. Luu, Mert \c{S}i\c{s}man, Alexey V. Dimov, Ceren Tozlu, Ilhami Kovanlikaya, Susan A. Gauthier, Thanh D. Nguyen, Yi Wang
Abstract: Quantitative susceptibility maps from magnetic resonance images can provide both prognostic and diagnostic information in multiple sclerosis, a neurodegenerative disease characterized by the formation of lesions in white matter brain tissue. In particular, susceptibility maps provide adequate contrast to distinguish between "rim" lesions, surrounded by deposited paramagnetic iron, and "non-rim" lesion types. These paramagnetic rim lesions (PRLs) are an emerging biomarker in multiple sclerosis. Much effort has been devoted to both detection and segmentation of such lesions to monitor longitudinal change. As paramagnetic rim lesions are rare, addressing this problem requires confronting the class imbalance between rim and non-rim lesions. We produce synthetic quantitative susceptibility maps of paramagnetic rim lesions and show that inclusion of such synthetic data improves classifier performance and provide a multi-channel extension to generate accompanying contrasts and probabilistic segmentation maps. We exploit the projection capability of our trained generative network to demonstrate a novel denoising approach that allows us to train on ambiguous rim cases and substantially increase the minority class. We show that both synthetic lesion synthesis and our proposed rim lesion label denoising method best approximate the unseen rim lesion distribution and improve detection in a clinically interpretable manner. We release our code and generated data at https://github.com/agr78/PRLx-GAN upon publication.
Authors: Meital Bojan, Sanketh Vedula, Advaith Maddipatla, Nadav Bojan Sellam, Federico Napoli, Paul Schanda, Alex M. Bronstein
Abstract: The local structure of a protein strongly impacts its function and interactions with other molecules. Therefore, a concise, informative representation of a local protein environment is essential for modeling and designing proteins and biomolecular interactions. However, these environments' extensive structural and chemical variability makes them challenging to model, and such representations remain under-explored. In this work, we propose a novel representation for a local protein environment derived from the intermediate features of atomistic foundation models (AFMs). We demonstrate that this embedding effectively captures both local structure (e.g., secondary motifs), and chemical features (e.g., amino-acid identity and protonation state). We further show that the AFM-derived representation space exhibits meaningful structure, enabling the construction of data-driven priors over the distribution of biomolecular environments. Finally, in the context of biomolecular NMR spectroscopy, we demonstrate that the proposed representations enable a first-of-its-kind physics-informed chemical shift predictor that achieves state-of-the-art accuracy. Our results demonstrate the surprising effectiveness of atomistic foundation models and their emergent representations for protein modeling beyond traditional molecular simulations. We believe this will open new lines of work in constructing effective functional representations for protein environments.
Authors: Jeonghyeok Do, Sungpyo Kim, Geunhyuk Youk, Jaehyup Lee, Munchurl Kim
Abstract: PAN-sharpening aims to fuse high-resolution panchromatic (PAN) images with low-resolution multi-spectral (MS) images to generate high-resolution multi-spectral (HRMS) outputs. However, cross-modality misalignment -- caused by sensor placement, acquisition timing, and resolution disparity -- induces a fundamental challenge. Conventional deep learning methods assume perfect pixel-wise alignment and rely on per-pixel reconstruction losses, leading to spectral distortion, double edges, and blurring when misalignment is present. To address this, we propose PAN-Crafter, a modality-consistent alignment framework that explicitly mitigates the misalignment gap between PAN and MS modalities. At its core, Modality-Adaptive Reconstruction (MARs) enables a single network to jointly reconstruct HRMS and PAN images, leveraging PAN's high-frequency details as auxiliary self-supervision. Additionally, we introduce Cross-Modality Alignment-Aware Attention (CM3A), a novel mechanism that bidirectionally aligns MS texture to PAN structure and vice versa, enabling adaptive feature refinement across modalities. Extensive experiments on multiple benchmark datasets demonstrate that our PAN-Crafter outperforms the most recent state-of-the-art method in all metrics, even with 50.11$\times$ faster inference time and 0.63$\times$ the memory size. Furthermore, it demonstrates strong generalization performance on unseen satellite datasets, showing its robustness across different conditions.
Authors: Mannmohan Muthuraman
Abstract: Dynamic Spectral Backpropagation (DSBP) enhances neural network training under resource constraints by projecting gradients onto principal eigenvectors, reducing complexity and promoting flat minima. Five extensions are proposed, dynamic spectral inference, spectral architecture optimization, spectral meta learning, spectral transfer regularization, and Lie algebra inspired dynamics, to address challenges in robustness, fewshot learning, and hardware efficiency. Supported by a third order stochastic differential equation (SDE) and a PAC Bayes limit, DSBP outperforms Sharpness Aware Minimization (SAM), Low Rank Adaptation (LoRA), and Model Agnostic Meta Learning (MAML) on CIFAR 10, Fashion MNIST, MedMNIST, and Tiny ImageNet, as demonstrated through extensive experiments and visualizations. Future work focuses on scalability, bias mitigation, and ethical considerations.
Authors: Han Bao, Qinying Wang, Zhi Chen, Qingming Li, Xuhong Zhang, Changjiang Li, Zonghui Wang, Shouling Ji, Wenzhi Chen
Abstract: Not Safe/Suitable for Work (NSFW) content is rampant on social networks and poses serious harm to citizens, especially minors. Current detection methods mainly rely on deep learning-based image recognition and classification. However, NSFW images are now presented in increasingly sophisticated ways, often using image details and complex semantics to obscure their true nature or attract more views. Although still understandable to humans, these images often evade existing detection methods, posing a significant threat. Further complicating the issue, varying regulations across platforms and regions create additional challenges for effective moderation, leading to detection bias and reduced accuracy. To address this, we propose VModA, a general and effective framework that adapts to diverse moderation rules and handles complex, semantically rich NSFW content across categories. Experimental results show that VModA significantly outperforms existing methods, achieving up to a 54.3% accuracy improvement across NSFW types, including those with complex semantics. Further experiments demonstrate that our method exhibits strong adaptability across categories, scenarios, and base VLMs. We also identified inconsistent and controversial label samples in public NSFW benchmark datasets, re-annotated them, and submitted corrections to the original maintainers. Two datasets have confirmed the updates so far. Additionally, we evaluate VModA in real-world scenarios to demonstrate its practical effectiveness.
Authors: Mingzhe Du, Luu Tuan Tuan, Yue Liu, Yuhao Qing, Dong Huang, Xinyi He, Qian Liu, Zejun Ma, See-kiong Ng
Abstract: Large Language Models (LLMs) generate functionally correct solutions but often fall short in code efficiency, a critical bottleneck for real-world deployment. In this paper, we introduce a novel test-time iterative optimization framework to address this, employing a closed-loop system where LLMs iteratively refine code based on empirical performance feedback from an execution sandbox. We explore three training strategies: Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Group Relative Policy Optimization~(GRPO). Experiments on our Venus dataset and the APPS benchmark show that SFT and DPO rapidly saturate in efficiency gains. In contrast, GRPO, using reinforcement learning (RL) with execution feedback, continuously optimizes code performance, significantly boosting both pass@1 (from 47% to 62%) and the likelihood of outperforming human submissions in efficiency (from 31% to 45%). Our work demonstrates effective test-time code efficiency improvement and critically reveals the power of RL in teaching LLMs to truly self-improve code efficiency.
Authors: Binyamin Manela, Sharon Gannot, Ethan Fetyaya
Abstract: Visual dubbing, the synchronization of facial movements with new speech, is crucial for making content accessible across different languages, enabling broader global reach. However, current methods face significant limitations. Existing approaches often generate talking faces, hindering seamless integration into original scenes, or employ inpainting techniques that discard vital visual information like partial occlusions and lighting variations. This work introduces EdiDub, a novel framework that reformulates visual dubbing as a content-aware editing task. EdiDub preserves the original video context by utilizing a specialized conditioning scheme to ensure faithful and accurate modifications rather than mere copying. On multiple benchmarks, including a challenging occluded-lip dataset, EdiDub significantly improves identity preservation and synchronization. Human evaluations further confirm its superiority, achieving higher synchronization and visual naturalness scores compared to the leading methods. These results demonstrate that our content-aware editing approach outperforms traditional generation or inpainting, particularly in maintaining complex visual elements while ensuring accurate lip synchronization.
Authors: Srishti Gupta, Daniele Angioni, Maura Pintor, Ambra Demontis, Lea Sch\"onherr, Battista Biggio, Fabio Roli
Abstract: Class-incremental learning (CIL) poses significant challenges in open-world scenarios, where models must not only learn new classes over time without forgetting previous ones but also handle inputs from unknown classes that a closed-set model would misclassify. Recent works address both issues by (i)~training multi-head models using the task-incremental learning framework, and (ii) predicting the task identity employing out-of-distribution (OOD) detectors. While effective, the latter mainly relies on joint training with a memory buffer of past data, raising concerns around privacy, scalability, and increased training time. In this paper, we present an in-depth analysis of post-hoc OOD detection methods and investigate their potential to eliminate the need for a memory buffer. We uncover that these methods, when applied appropriately at inference time, can serve as a strong substitute for buffer-based OOD detection. We show that this buffer-free approach achieves comparable or superior performance to buffer-based methods both in terms of class-incremental learning and the rejection of unknown samples. Experimental results on CIFAR-10, CIFAR-100 and Tiny ImageNet datasets support our findings, offering new insights into the design of efficient and privacy-preserving CIL systems for open-world settings.
Authors: Gaspard Oliviers, Mufeng Tang, Rafal Bogacz
Abstract: Predictive coding (PC) is an influential computational model of visual learning and inference in the brain. Classical PC was proposed as a top-down generative model, where the brain actively predicts upcoming visual inputs, and inference minimises the prediction errors. Recent studies have also shown that PC can be formulated as a discriminative model, where sensory inputs predict neural activities in a feedforward manner. However, experimental evidence suggests that the brain employs both generative and discriminative inference, while unidirectional PC models show degraded performance in tasks requiring bidirectional processing. In this work, we propose bidirectional PC (bPC), a PC model that incorporates both generative and discriminative inference while maintaining a biologically plausible circuit implementation. We show that bPC matches or outperforms unidirectional models in their specialised generative or discriminative tasks, by developing an energy landscape that simultaneously suits both tasks. We also demonstrate bPC's superior performance in two biologically relevant tasks including multimodal learning and inference with missing information, suggesting that bPC resembles biological visual inference more closely.
Authors: Danilo Ribeiro, Thayssa Rocha, Gustavo Pinto, Bruno Cartaxo, Marcelo Amaral, Nicole Davila, Ana Camargo
Abstract: Artificial Intelligence (AI) governance is the practice of establishing frameworks, policies, and procedures to ensure the responsible, ethical, and safe development and deployment of AI systems. Although AI governance is a core pillar of Responsible AI, current literature still lacks synthesis across such governance frameworks and practices. Objective: To identify which frameworks, principles, mechanisms, and stakeholder roles are emphasized in secondary literature on AI governance. Method: We conducted a rapid tertiary review of nine peer-reviewed secondary studies from IEEE and ACM (20202024), using structured inclusion criteria and thematic semantic synthesis. Results: The most cited frameworks include the EU AI Act and NIST RMF; transparency and accountability are the most common principles. Few reviews detail actionable governance mechanisms or stakeholder strategies. Conclusion: The review consolidates key directions in AI governance and highlights gaps in empirical validation and inclusivity. Findings inform both academic inquiry and practical adoption in organizations.
Authors: Linghao Zhang, Shilin He, Chaoyun Zhang, Yu Kang, Bowen Li, Chengxing Xie, Junhao Wang, Maoquan Wang, Yufan Huang, Shengyu Fu, Elsie Nallipogu, Qingwei Lin, Yingnong Dang, Saravan Rajmohan, Dongmei Zhang
Abstract: The issue-resolving task, where a model generates patches to fix real-world bugs, has emerged as a critical benchmark for evaluating the capabilities of large language models (LLMs). While SWE-bench and its variants have become standard in this domain, they suffer from key limitations: they have not been updated since their initial releases, cover a narrow set of repositories, and depend heavily on manual effort for instance construction and environment setup. These factors hinder scalability and introduce risks of overfitting and data contamination. In this work, we present \textbf{SWE-bench-Live}, a \textit{live-updatable} benchmark designed to overcome these challenges. Our initial release consists of 1,319 tasks derived from real GitHub issues created since 2024, spanning 93 repositories. Each task is accompanied by a dedicated Docker image to ensure reproducible execution. Central to our benchmark is \method, an automated curation pipeline that streamlines the entire process from instance creation to environment setup, removing manual bottlenecks and enabling scalability and continuous updates. We evaluate a range of state-of-the-art agent frameworks and LLMs on SWE-bench-Live, revealing a substantial performance gap compared to static benchmarks like SWE-bench, even under controlled evaluation conditions. To better understand this discrepancy, we perform detailed analyses across repository origin, issue recency, and task difficulty. By providing a fresh, diverse, and executable benchmark grounded in live repository activity, SWE-bench-Live facilitates rigorous, contamination-resistant evaluation of LLMs and agents in dynamic, real-world software development settings.
Authors: Tobias Lindenbauer, Georg Groh, Hinrich Sch\"utze
Abstract: We introduce CTIM-Rover, an AI agent for Software Engineering (SE) built on top of AutoCodeRover (Zhang et al., 2024) that extends agentic reasoning frameworks with an episodic memory, more specifically, a general and repository-level Cross-Task-Instance Memory (CTIM). While existing open-source SE agents mostly rely on ReAct (Yao et al., 2023b), Reflexion (Shinn et al., 2023), or Code-Act (Wang et al., 2024), all of these reasoning and planning frameworks inefficiently discard their long-term memory after a single task instance. As repository-level understanding is pivotal for identifying all locations requiring a patch for fixing a bug, we hypothesize that SE is particularly well positioned to benefit from CTIM. For this, we build on the Experiential Learning (EL) approach ExpeL (Zhao et al., 2024), proposing a Mixture-Of-Experts (MoEs) inspired approach to create both a general-purpose and repository-level CTIM. We find that CTIM-Rover does not outperform AutoCodeRover in any configuration and thus conclude that neither ExpeL nor DoT-Bank (Lingam et al., 2024) scale to real-world SE problems. Our analysis indicates noise introduced by distracting CTIM items or exemplar trajectories as the likely source of the performance degradation.
Authors: Yinuo Wang, Mining Tan, Wenjun Zou, Haotian Lin, Xujie Song, Wenxuan Wang, Tong Liu, Likun Wang, Guojian Zhan, Tianze Zhu, Shiqi Liu, Jingliang Duan, Shengbo Eben Li
Abstract: Due to their expressive capacity, diffusion models have shown great promise in offline RL and imitation learning. Diffusion Actor-Critic with Entropy Regulator (DACER) extended this capability to online RL by using the reverse diffusion process as a policy approximator, trained end-to-end with policy gradient methods, achieving strong performance. However, this comes at the cost of requiring many diffusion steps, which significantly hampers training efficiency, while directly reducing the steps leads to noticeable performance degradation. Critically, the lack of inference efficiency becomes a significant bottleneck for applying diffusion policies in real-time online RL settings. To improve training and inference efficiency while maintaining or even enhancing performance, we propose a Q-gradient field objective as an auxiliary optimization target to guide the denoising process at each diffusion step. Nonetheless, we observe that the independence of the Q-gradient field from the diffusion time step negatively impacts the performance of the diffusion policy. To address this, we introduce a temporal weighting mechanism that enables the model to efficiently eliminate large-scale noise in the early stages and refine actions in the later stages. Experimental results on MuJoCo benchmarks and several multimodal tasks demonstrate that the DACER2 algorithm achieves state-of-the-art performance in most MuJoCo control tasks with only five diffusion steps, while also exhibiting stronger multimodality compared to DACER.
Authors: Antonio Ferrara, Andrea Pugnana, Francesco Bonchi, Salvatore Ruggieri
Abstract: Ranking systems influence decision-making in high-stakes domains like health, education, and employment, where they can have substantial economic and social impacts. This makes the integration of safety mechanisms essential. One such mechanism is $\textit{abstention}$, which enables algorithmic decision-making system to defer uncertain or low-confidence decisions to human experts. While abstention have been predominantly explored in the context of classification tasks, its application to other machine learning paradigms remains underexplored. In this paper, we introduce a novel method for abstention in pairwise learning-to-rank tasks. Our approach is based on thresholding the ranker's conditional risk: the system abstains from making a decision when the estimated risk exceeds a predefined threshold. Our contributions are threefold: a theoretical characterization of the optimal abstention strategy, a model-agnostic, plug-in algorithm for constructing abstaining ranking models, and a comprehensive empirical evaluations across multiple datasets, demonstrating the effectiveness of our approach.
Authors: Runmin Jiang, Genpei Zhang, Yuntian Yang, Siqi Wu, Yuheng Zhang, Wanyue Feng, Yizhou Zhao, Xi Xiao, Xiao Wang, Tianyang Wang, Xingjian Li, Min Xu
Abstract: Cryo-electron microscopy (cryo-EM) offers near-atomic resolution imaging of macromolecules, but developing robust models for downstream analysis is hindered by the scarcity of high-quality annotated data. While synthetic data generation has emerged as a potential solution, existing methods often fail to capture both the structural diversity of biological specimens and the complex, spatially varying noise inherent in cryo-EM imaging. To overcome these limitations, we propose CryoCCD, a synthesis framework that integrates biophysical modeling with generative techniques. Specifically, CryoCCD produces multi-scale cryo-EM micrographs that reflect realistic biophysical variability through compositional heterogeneity, cellular context, and physics-informed imaging. To generate realistic noise, we employ a conditional diffusion model, enhanced by cycle consistency to preserve structural fidelity and mask-aware contrastive learning to capture spatially adaptive noise patterns. Extensive experiments show that CryoCCD generates structurally accurate micrographs and enhances performance in downstream tasks, outperforming state-of-the-art baselines in both particle picking and reconstruction.
Authors: Yanbin Wang, Xingyu Chen, Yumiao Wang, Xiang Wang, Chuanfei Zang, Guolong Cui, Jiahuan Liu
Abstract: We propose the LCB-CV-UNet to tackle performance degradation caused by High Dynamic Range (HDR) radar signals. Initially, a hardware-efficient, plug-and-play module named Logarithmic Connect Block (LCB) is proposed as a phase coherence preserving solution to address the inherent challenges in handling HDR features. Then, we propose the Dual Hybrid Dataset Construction method to generate a semi-synthetic dataset, approximating typical HDR signal scenarios with adjustable target distributions. Simulation results show about 1% total detection probability improvement with under 0.9% computational complexity added compared with the baseline. Furthermore, it excels 5% over the baseline at the range in 11-13 dB signal-to-noise ratio typical for urban targets. Finally, the real experiment validates the practicality of our model.
Authors: Shibbir Ahmed, Shahnewaz Karim Sakib, Anindya Bijoy Das
Abstract: This study presents a multimodal AI framework designed for precisely classifying medical diagnostic images. Utilizing publicly available datasets, the proposed system compares the strengths of convolutional neural networks (CNNs) and different large language models (LLMs). This in-depth comparative analysis highlights key differences in diagnostic performance, execution efficiency, and environmental impacts. Model evaluation was based on accuracy, F1-score, average execution time, average energy consumption, and estimated $CO_2$ emission. The findings indicate that although CNN-based models can outperform various multimodal techniques that incorporate both images and contextual information, applying additional filtering on top of LLMs can lead to substantial performance gains. These findings highlight the transformative potential of multimodal AI systems to enhance the reliability, efficiency, and scalability of medical diagnostics in clinical settings.
Authors: Rebecca Ramnauth, Dra\v{z}en Br\v{s}\v{c}i\'c, Brian Scassellati
Abstract: From dating to job interviews, making new friends or simply chatting with the cashier at checkout, engaging in small talk is a vital, everyday social skill. For adults with Autism Spectrum Disorder (ASD), small talk can be particularly challenging, yet it is essential for social integration, building relationships, and accessing professional opportunities. In this study, we present our development and evaluation of an in-home autonomous robot system that allows users to practice small talk. Results from the week-long study show that adults with ASD enjoyed the training, made notable progress in initiating conversations and improving eye contact, and viewed the system as a valuable tool for enhancing their conversational skills.
Authors: Shifeng Xie, Aref Einizade, Jhony H. Giraldo
Abstract: Graph Representation Learning (GRL) is a fundamental task in machine learning, aiming to encode high-dimensional graph-structured data into low-dimensional vectors. Self-Supervised Learning (SSL) methods are widely used in GRL because they can avoid expensive human annotation. In this work, we propose a novel Subgraph Gaussian Embedding Contrast (SubGEC) method. Our approach introduces a subgraph Gaussian embedding module, which adaptively maps subgraphs to a structured Gaussian space, ensuring the preservation of input subgraph characteristics while generating subgraphs with a controlled distribution. We then employ optimal transport distances, more precisely the Wasserstein and Gromov-Wasserstein distances, to effectively measure the similarity between subgraphs, enhancing the robustness of the contrastive learning process. Extensive experiments across multiple benchmarks demonstrate that \method~outperforms or presents competitive performance against state-of-the-art approaches. Our findings provide insights into the design of SSL methods for GRL, emphasizing the importance of the distribution of the generated contrastive pairs.
Authors: Nawar Turk, Eeham Khan, Leila Kosseim
Abstract: This paper presents our approach to the SemEval-2025 Task~6 (PromiseEval), which focuses on verifying promises in corporate ESG (Environmental, Social, and Governance) reports. We explore three model architectures to address the four subtasks of promise identification, supporting evidence assessment, clarity evaluation, and verification timing. Our first model utilizes ESG-BERT with task-specific classifier heads, while our second model enhances this architecture with linguistic features tailored for each subtask. Our third approach implements a combined subtask model with attention-based sequence pooling, transformer representations augmented with document metadata, and multi-objective learning. Experiments on the English portion of the ML-Promise dataset demonstrate progressive improvement across our models, with our combined subtask approach achieving a leaderboard score of 0.5268, outperforming the provided baseline of 0.5227. Our work highlights the effectiveness of linguistic feature extraction, attention pooling, and multi-objective learning in promise verification tasks, despite challenges posed by class imbalance and limited training data.
Authors: Hayden Moore, Sirui Qi, Ninad Hogade, Dejan Milojicic, Cullen Bash, Sudeep Pasricha
Abstract: In recent years, Large Language Models (LLM) such as ChatGPT, CoPilot, and Gemini have been widely adopted in different areas. As the use of LLMs continues to grow, many efforts have focused on reducing the massive training overheads of these models. But it is the environmental impact of handling user requests to LLMs that is increasingly becoming a concern. Recent studies estimate that the costs of operating LLMs in their inference phase can exceed training costs by 25x per year. As LLMs are queried incessantly, the cumulative carbon footprint for the operational phase has been shown to far exceed the footprint during the training phase. Further, estimates indicate that 500 ml of fresh water is expended for every 20-50 requests to LLMs during inference. To address these important sustainability issues with LLMs, we propose a novel framework called SLIT to co-optimize LLM quality of service (time-to-first token), carbon emissions, water usage, and energy costs. The framework utilizes a machine learning (ML) based metaheuristic to enhance the sustainability of LLM hosting across geo-distributed cloud datacenters. Such a framework will become increasingly vital as LLMs proliferate.
Authors: Yiran Guo, Lijie Xu, Jie Liu, Dan Ye, Shuang Qiu
Abstract: Enhancing the reasoning capabilities of large language models effectively using reinforcement learning (RL) remains a crucial challenge. Existing approaches primarily adopt two contrasting advantage estimation granularities: Token-level methods (e.g., PPO) aim to provide the fine-grained advantage signals but suffer from inaccurate estimation due to difficulties in training an accurate critic model. On the other extreme, trajectory-level methods (e.g., GRPO) solely rely on a coarse-grained advantage signal from the final reward, leading to imprecise credit assignment. To address these limitations, we propose Segment Policy Optimization (SPO), a novel RL framework that leverages segment-level advantage estimation at an intermediate granularity, achieving a better balance by offering more precise credit assignment than trajectory-level methods and requiring fewer estimation points than token-level methods, enabling accurate advantage estimation based on Monte Carlo (MC) without a critic model. SPO features three components with novel strategies: (1) flexible segment partition; (2) accurate segment advantage estimation; and (3) policy optimization using segment advantages, including a novel probability-mask strategy. We further instantiate SPO for two specific scenarios: (1) SPO-chain for short chain-of-thought (CoT), featuring novel cutpoint-based partition and chain-based advantage estimation, achieving $6$-$12$ percentage point improvements in accuracy over PPO and GRPO on GSM8K. (2) SPO-tree for long CoT, featuring novel tree-based advantage estimation, which significantly reduces the cost of MC estimation, achieving $7$-$11$ percentage point improvements over GRPO on MATH500 under 2K and 4K context evaluation. We make our code publicly available at https://github.com/AIFrameResearch/SPO.
Authors: Jane Cleland-Huang, Pedro Antonio Alarcon Granadeno, Arturo Miguel Russell Bernal, Demetrius Hernandez, Michael Murphy, Maureen Petterson, Walter Scheirer
Abstract: Small Uncrewed Aerial Systems (sUAS) are increasingly deployed as autonomous swarms in search-and-rescue and other disaster-response scenarios. In these settings, they use computer vision (CV) to detect objects of interest and autonomously adapt their missions. However, traditional CV systems often struggle to recognize unfamiliar objects in open-world environments or to infer their relevance for mission planning. To address this, we incorporate large language models (LLMs) to reason about detected objects and their implications. While LLMs can offer valuable insights, they are also prone to hallucinations and may produce incorrect, misleading, or unsafe recommendations. To ensure safe and sensible decision-making under uncertainty, high-level decisions must be governed by cognitive guardrails. This article presents the design, simulation, and real-world integration of these guardrails for sUAS swarms in search-and-rescue missions.
Authors: Ramit Aditya, Razvan Bunescu, Smita Nannaware, Erfan Al-Hossami
Abstract: A restaurant dinner or a hotel stay may lead to memorable experiences when guests encounter unexpected aspects that also match their interests. For example, an origami-making station in the waiting area of a restaurant may be both surprising and enjoyable for a customer who is passionate about paper crafts. Similarly, an exhibit of 18th century harpsichords would be atypical for a hotel lobby and likely pique the interest of a guest who has a passion for Baroque music. Motivated by this insight, in this paper we introduce the new task of engineering serendipity through recommendations of items with atypical aspects. We describe an LLM-based system pipeline that extracts atypical aspects from item reviews, then estimates and aggregates their user-specific utility in a measure of serendipity potential that is used to rerank a list of items recommended to the user. To facilitate system development and evaluation, we introduce a dataset of Yelp reviews that are manually annotated with atypical aspects and a dataset of artificially generated user profiles, together with crowdsourced annotations of user-aspect utility values. Furthermore, we introduce a custom procedure for dynamic selection of in-context learning examples, which is shown to improve LLM-based judgments of atypicality and utility. Experimental evaluations show that serendipity-based rankings generated by the system are highly correlated with ground truth rankings for which serendipity scores are computed from manual annotations of atypical aspects and their user-dependent utility. Overall, we hope that the new recommendation task and the associated system presented in this paper catalyze further research into recommendation approaches that go beyond accuracy in their pursuit of enhanced user satisfaction. The datasets and the code are made publicly available at https://github.com/ramituncc49er/ATARS .
Authors: Sumbal Malik, Majid Khonji, Khaled Elbassioni, Jorge Dias
Abstract: The rapid growth of e-commerce and the increasing demand for timely, cost-effective last-mile delivery have increased interest in collaborative logistics. This research introduces a novel collaborative synchronized multi-platform vehicle routing problem with drones and robots (VRP-DR), where a fleet of $\mathcal{M}$ trucks, $\mathcal{N}$ drones and $\mathcal{K}$ robots, cooperatively delivers parcels. Trucks serve as mobile platforms, enabling the launching, retrieving, and en-route charging of drones and robots, thereby addressing critical limitations such as restricted payload capacities, limited range, and battery constraints. The VRP-DR incorporates five realistic features: (1) multi-visit service per trip, (2) multi-trip operations, (3) flexible docking, allowing returns to the same or different trucks (4) cyclic and acyclic operations, enabling return to the same or different nodes; and (5) en-route charging, enabling drones and robots to recharge while being transported on the truck, maximizing operational efficiency by utilizing idle transit time. The VRP-DR is formulated as a mixed-integer linear program (MILP) to minimize both operational costs and makespan. To overcome the computational challenges of solving large-scale instances, a scalable heuristic algorithm, FINDER (Flexible INtegrated Delivery with Energy Recharge), is developed, to provide efficient, near-optimal solutions. Numerical experiments across various instance sizes evaluate the performance of the MILP and heuristic approaches in terms of solution quality and computation time. The results demonstrate significant time savings of the combined delivery mode over the truck-only mode and substantial cost reductions from enabling multi-visits. The study also provides insights into the effects of en-route charging, docking flexibility, drone count, speed, and payload capacity on system performance.
Authors: Zifu Wang, Junyi Zhu, Bo Tang, Zhiyu Li, Feiyu Xiong, Jiaqian Yu, Matthew B. Blaschko
Abstract: The application of rule-based reinforcement learning (RL) to multimodal large language models (MLLMs) introduces unique challenges and potential deviations from findings in text-only domains, particularly for perception-heavy tasks. This paper provides a comprehensive study of rule-based visual RL using jigsaw puzzles as a structured experimental framework, revealing several key findings. \textit{Firstly,} we find that MLLMs, initially performing near to random guessing on simple puzzles, achieve near-perfect accuracy and generalize to complex, unseen configurations through fine-tuning. \textit{Secondly,} training on jigsaw puzzles can induce generalization to other visual tasks, with effectiveness tied to specific task configurations. \textit{Thirdly,} MLLMs can learn and generalize with or without explicit reasoning, though open-source models often favor direct answering. Consequently, even when trained for step-by-step reasoning, they can ignore the thinking process in deriving the final answer. \textit{Fourthly,} we observe that complex reasoning patterns appear to be pre-existing rather than emergent, with their frequency increasing alongside training and task difficulty. \textit{Finally,} our results demonstrate that RL exhibits more effective generalization than Supervised Fine-Tuning (SFT), and an initial SFT cold start phase can hinder subsequent RL optimization. Although these observations are based on jigsaw puzzles and may vary across other visual tasks, this research contributes a valuable piece of jigsaw to the larger puzzle of collective understanding rule-based visual RL and its potential in multimodal learning. The code is available at: \href{https://github.com/zifuwanggg/Jigsaw-R1}{https://github.com/zifuwanggg/Jigsaw-R1}.
URLs: https://github.com/zifuwanggg/Jigsaw-R1, https://github.com/zifuwanggg/Jigsaw-R1
Authors: Youssef Mohamed, Noran Mohamed, Khaled Abouhashad, Feilong Tang, Sara Atito, Shoaib Jameel, Imran Razzak, Ahmed B. Zaky
Abstract: While Multi-Task Learning (MTL) offers inherent advantages in complex domains such as medical imaging by enabling shared representation learning, effectively balancing task contributions remains a significant challenge. This paper addresses this critical issue by introducing DeepChest, a novel, computationally efficient and effective dynamic task-weighting framework specifically designed for multi-label chest X-ray (CXR) classification. Unlike existing heuristic or gradient-based methods that often incur substantial overhead, DeepChest leverages a performance-driven weighting mechanism based on effective analysis of task-specific loss trends. Given a network architecture (e.g., ResNet18), our model-agnostic approach adaptively adjusts task importance without requiring gradient access, thereby significantly reducing memory usage and achieving a threefold increase in training speed. It can be easily applied to improve various state-of-the-art methods. Extensive experiments on a large-scale CXR dataset demonstrate that DeepChest not only outperforms state-of-the-art MTL methods by 7% in overall accuracy but also yields substantial reductions in individual task losses, indicating improved generalization and effective mitigation of negative transfer. The efficiency and performance gains of DeepChest pave the way for more practical and robust deployment of deep learning in critical medical diagnostic applications. The code is publicly available at https://github.com/youssefkhalil320/DeepChest-MTL
Authors: Guangtao Zeng, Maohao Shen, Delin Chen, Zhenting Qi, Subhro Das, Dan Gutfreund, David Cox, Gregory Wornell, Wei Lu, Zhang-Wei Hong, Chuang Gan
Abstract: Language models (LMs) perform well on standardized coding benchmarks but struggle with real-world software engineering tasks such as resolving GitHub issues in SWE-Bench, especially when model parameters are less than 100B. While smaller models are preferable in practice due to their lower computational cost, improving their performance remains challenging. Existing approaches primarily rely on supervised fine-tuning (SFT) with high-quality data, which is expensive to curate at scale. An alternative is test-time scaling: generating multiple outputs, scoring them using a verifier, and selecting the best one. Although effective, this strategy often requires excessive sampling and costly scoring, limiting its practical application. We propose Evolutionary Test-Time Scaling (EvoScale), a sample-efficient method that treats generation as an evolutionary process. By iteratively refining outputs via selection and mutation, EvoScale shifts the output distribution toward higher-scoring regions, reducing the number of samples needed to find correct solutions. To reduce the overhead from repeatedly sampling and selection, we train the model to self-evolve using reinforcement learning (RL). Rather than relying on external verifiers at inference time, the model learns to self-improve the scores of its own generations across iterations. Evaluated on SWE-Bench-Verified, EvoScale enables our 32B model, Satori-SWE-32B, to match or exceed the performance of models with over 100B parameters while using a few samples. Code, data, and models will be fully open-sourced.
Authors: Chenhao Zheng, Jieyu Zhang, Mohammadreza Salehi, Ziqi Gao, Vishnu Iyengar, Norimasa Kobori, Quan Kong, Ranjay Krishna
Abstract: Effective video tokenization is critical for scaling transformer models for long videos. Current approaches tokenize videos using space-time patches, leading to excessive tokens and computational inefficiencies. The best token reduction strategies degrade performance and barely reduce the number of tokens when the camera moves. We introduce grounded video tokenization, a paradigm that organizes tokens based on panoptic sub-object trajectories rather than fixed patches. Our method aligns with fundamental perceptual principles, ensuring that tokenization reflects scene complexity rather than video duration. We propose TrajViT, a video encoder that extracts object trajectories and converts them into semantically meaningful tokens, significantly reducing redundancy while maintaining temporal coherence. Trained with contrastive learning, TrajViT significantly outperforms space-time ViT (ViT3D) across multiple video understanding benchmarks, e.g., TrajViT outperforms ViT3D by a large margin of 6% top-5 recall in average at video-text retrieval task with 10x token deduction. We also show TrajViT as a stronger model than ViT3D for being the video encoder for modern VideoLLM, obtaining an average of 5.2% performance improvement across 6 VideoQA benchmarks while having 4x faster training time and 18x less inference FLOPs. TrajViT is the first efficient encoder to consistently outperform ViT3D across diverse video analysis tasks, making it a robust and scalable solution.
Authors: Giacomo Bergami, Emma Packer, Kirsty Scott, Silvia Del Din
Abstract: This paper offers a hybrid explainable temporal data processing pipeline, DataFul Explainable MultivariatE coRrelatIonal Temporal Artificial inTElligence (EMeriTAte+DF), bridging numerical-driven temporal data classification with an event-based one through verified artificial intelligence principles, enabling human-explainable results. This was possible through a preliminary a posteriori explainable phase describing the numerical input data in terms of concurrent constituents with numerical payloads. This further required extending the event-based literature to design specification mining algorithms supporting concurrent constituents. Our previous and current solutions outperform state-of-the-art solutions for multivariate time series classifications, thus showcasing the effectiveness of the proposed methodology.
Authors: Jiaxin Bai, Wei Fan, Qi Hu, Qing Zong, Chunyang Li, Hong Ting Tsang, Hongyu Luo, Yauwai Yim, Haoyu Huang, Xiao Zhou, Feng Qin, Tianshi Zheng, Xi Peng, Xin Yao, Huiwen Yang, Leijie Wu, Yi Ji, Gong Zhang, Renhai Chen, Yangqiu Song
Abstract: We present AutoSchemaKG, a framework for fully autonomous knowledge graph construction that eliminates the need for predefined schemas. Our system leverages large language models to simultaneously extract knowledge triples and induce comprehensive schemas directly from text, modeling both entities and events while employing conceptualization to organize instances into semantic categories. Processing over 50 million documents, we construct ATLAS (Automated Triple Linking And Schema induction), a family of knowledge graphs with 900+ million nodes and 5.9 billion edges. This approach outperforms state-of-the-art baselines on multi-hop QA tasks and enhances LLM factuality. Notably, our schema induction achieves 95\% semantic alignment with human-crafted schemas with zero manual intervention, demonstrating that billion-scale knowledge graphs with dynamically induced schemas can effectively complement parametric knowledge in large language models.
Authors: Boning Zhao
Abstract: Assessing student depression in sensitive environments like special education is challenging. Standardized questionnaires may not fully reflect students' true situations. Furthermore, automated methods often falter with rich student narratives, lacking the crucial, individualized insights stemming from teachers' empathetic connections with students. Existing methods often fail to address this ambiguity or effectively integrate educator understanding. To address these limitations by fostering a synergistic human-AI collaboration, this paper introduces Human Empathy as Encoder (HEAE), a novel, human-centered AI framework for transparent and socially responsible depression severity assessment. Our approach uniquely integrates student narrative text with a teacher-derived, 9-dimensional "Empathy Vector" (EV), its dimensions guided by the PHQ-9 framework,to explicitly translate tacit empathetic insight into a structured AI input enhancing rather than replacing human judgment. Rigorous experiments optimized the multimodal fusion, text representation, and classification architecture, achieving 82.74% accuracy for 7-level severity classification. This work demonstrates a path toward more responsible and ethical affective computing by structurally embedding human empathy
Authors: Dashti A. Ali, Richard K. G. Do, William R. Jarnagin, Aras T. Asaad, Amber L. Simpson
Abstract: In medical image analysis, feature engineering plays an important role in the design and performance of machine learning models. Persistent homology (PH), from the field of topological data analysis (TDA), demonstrates robustness and stability to data perturbations and addresses the limitation from traditional feature extraction approaches where a small change in input results in a large change in feature representation. Using PH, we store persistent topological and geometrical features in the form of the persistence barcode whereby large bars represent global topological features and small bars encapsulate geometrical information of the data. When multiple barcodes are computed from 2D or 3D medical images, two approaches can be used to construct the final topological feature vector in each dimension: aggregating persistence barcodes followed by featurization or concatenating topological feature vectors derived from each barcode. In this study, we conduct a comprehensive analysis across diverse medical imaging datasets to compare the effects of the two aforementioned approaches on the performance of classification models. The results of this analysis indicate that feature concatenation preserves detailed topological information from individual barcodes, yields better classification performance and is therefore a preferred approach when conducting similar experiments.
Authors: Manuel Costa, Boris K\"opf, Aashish Kolluri, Andrew Paverd, Mark Russinovich, Ahmed Salem, Shruti Tople, Lukas Wutschitz, Santiago Zanella-B\'eguelin
Abstract: As AI agents become increasingly autonomous and capable, ensuring their security against vulnerabilities such as prompt injection becomes critical. This paper explores the use of information-flow control (IFC) to provide security guarantees for AI agents. We present a formal model to reason about the security and expressiveness of agent planners. Using this model, we characterize the class of properties enforceable by dynamic taint-tracking and construct a taxonomy of tasks to evaluate security and utility trade-offs of planner designs. Informed by this exploration, we present Fides, a planner that tracks confidentiality and integrity labels, deterministically enforces security policies, and introduces novel primitives for selectively hiding information. Its evaluation in AgentDojo demonstrates that this approach broadens the range of tasks that can be securely accomplished. A tutorial to walk readers through the the concepts introduced in the paper can be found at https://github.com/microsoft/fides
Authors: Peter David Fagan
Abstract: This work introduces a novel framework for secure and privacy-preserving neural network inference based on keyed chaotic dynamical transformations. The proposed method applies a deterministic, cryptographically seeded chaotic system to tensors, producing non-invertible, user-specific transformations that enable authenticated inference, tensor-level watermarking, and data attribution. This framework offers a scalable and lightweight alternative to conventional cryptographic techniques, and establishes a new direction for tensor-level security in AI systems.
Authors: Hongxiang Zhang, Hao Chen, Tianyi Zhang, Muhao Chen
Abstract: Recent decoding methods improve the factuality of large language models~(LLMs) by refining how the next token is selected during generation. These methods typically operate at the token level, leveraging internal representations to suppress superficial patterns. Nevertheless, LLMs remain prone to hallucinations, especially over longer contexts. In this paper, we propose Active Layer-Contrastive Decoding (ActLCD), a novel decoding strategy that actively decides when to apply contrasting layers during generation. By casting decoding as a sequential decision-making problem, ActLCD employs a reinforcement learning policy guided by a reward-aware classifier to optimize factuality beyond the token level. Our experiments demonstrate that ActLCD surpasses state-of-the-art methods across five benchmarks, showcasing its effectiveness in mitigating hallucinations in diverse generation scenarios.
Authors: Manish Shetty, Naman Jain, Jinjian Liu, Vijay Kethanaboyina, Koushik Sen, Ion Stoica
Abstract: Developing high-performance software is a complex task that requires specialized expertise. We introduce GSO, a benchmark for evaluating language models' capabilities in developing high-performance software. We develop an automated pipeline that generates and executes performance tests to analyze repository commit histories to identify 102 challenging optimization tasks across 10 codebases, spanning diverse domains and programming languages. An agent is provided with a codebase and performance test as a precise specification, and tasked to improve the runtime efficiency, which is measured against the expert developer optimization. Our quantitative evaluation reveals that leading SWE-Agents struggle significantly, achieving less than 5% success rate, with limited improvements even with inference-time scaling. Our qualitative analysis identifies key failure modes, including difficulties with low-level languages, practicing lazy optimization strategies, and challenges in accurately localizing bottlenecks. We release the code and artifacts of our benchmark along with agent trajectories to enable future research.
Authors: Tingyu Song, Tongyan Hu, Guo Gan, Yilun Zhao
Abstract: MLLMs have been widely studied for video question answering recently. However, most existing assessments focus on natural videos, overlooking synthetic videos, such as AI-generated content (AIGC). Meanwhile, some works in video generation rely on MLLMs to evaluate the quality of generated videos, but the capabilities of MLLMs on interpreting AIGC videos remain largely underexplored. To address this, we propose a new benchmark, VF-Eval, which introduces four tasks-coherence validation, error awareness, error type detection, and reasoning evaluation-to comprehensively evaluate the abilities of MLLMs on AIGC videos. We evaluate 13 frontier MLLMs on VF-Eval and find that even the best-performing model, GPT-4.1, struggles to achieve consistently good performance across all tasks. This highlights the challenging nature of our benchmark. Additionally, to investigate the practical applications of VF-Eval in improving video generation, we conduct an experiment, RePrompt, demonstrating that aligning MLLMs more closely with human feedback can benefit video generation.
Authors: Mohamad Alansari, Sajid Javed, Iyyakutti Iyappan Ganapathi, Sara Alansari, Muzammal Naseer
Abstract: VOT remains a fundamental yet challenging task in computer vision due to dynamic appearance changes, occlusions, and background clutter. Traditional trackers, relying primarily on visual cues, often struggle in such complex scenarios. Recent advancements in VLMs have shown promise in semantic understanding for tasks like open-vocabulary detection and image captioning, suggesting their potential for VOT. However, the direct application of VLMs to VOT is hindered by critical limitations: the absence of a rich and comprehensive textual representation that semantically captures the target object's nuances, limiting the effective use of language information; inefficient fusion mechanisms that fail to optimally integrate visual and textual features, preventing a holistic understanding of the target; and a lack of temporal modeling of the target's evolving appearance in the language domain, leading to a disconnect between the initial description and the object's subsequent visual changes. To bridge these gaps and unlock the full potential of VLMs for VOT, we propose CLDTracker, a novel Comprehensive Language Description framework for robust visual Tracking. Our tracker introduces a dual-branch architecture consisting of a textual and a visual branch. In the textual branch, we construct a rich bag of textual descriptions derived by harnessing the powerful VLMs such as CLIP and GPT-4V, enriched with semantic and contextual cues to address the lack of rich textual representation. Experiments on six standard VOT benchmarks demonstrate that CLDTracker achieves SOTA performance, validating the effectiveness of leveraging robust and temporally-adaptive vision-language representations for tracking. Code and models are publicly available at: https://github.com/HamadYA/CLDTracker
Authors: Utku Demir, Yalin E. Sagduyu, Tugba Erpek, Hossein Jafari, Sastry Kompella, Mengran Xue
Abstract: In connected and autonomous vehicles, machine learning for safety message classification has become critical for detecting malicious or anomalous behavior. However, conventional approaches that rely on centralized data collection or purely local training face limitations due to the large scale, high mobility, and heterogeneous data distributions inherent in inter-vehicle networks. To overcome these challenges, this paper explores Distributed Federated Learning (DFL), whereby vehicles collaboratively train deep learning models by exchanging model updates among one-hop neighbors and propagating models over multiple hops. Using the Vehicular Reference Misbehavior (VeReMi) Extension Dataset, we show that DFL can significantly improve classification accuracy across all vehicles compared to learning strictly with local data. Notably, vehicles with low individual accuracy see substantial accuracy gains through DFL, illustrating the benefit of knowledge sharing across the network. We further show that local training data size and time-varying network connectivity correlate strongly with the model's overall accuracy. We investigate DFL's resilience and vulnerabilities under attacks in multiple domains, namely wireless jamming and training data poisoning attacks. Our results reveal important insights into the vulnerabilities of DFL when confronted with multi-domain attacks, underlining the need for more robust strategies to secure DFL in vehicular networks.
Authors: Dionysis Christopoulos, Sotiris Spanos, Eirini Baltzi, Valsamis Ntouskos, Konstantinos Karantzalos
Abstract: We introduce SLIMP (Skin Lesion Image-Metadata Pre-training) for learning rich representations of skin lesions through a novel nested contrastive learning approach that captures complex relationships between images and metadata. Melanoma detection and skin lesion classification based solely on images, pose significant challenges due to large variations in imaging conditions (lighting, color, resolution, distance, etc.) and lack of clinical and phenotypical context. Clinicians typically follow a holistic approach for assessing the risk level of the patient and for deciding which lesions may be malignant and need to be excised, by considering the patient's medical history as well as the appearance of other lesions of the patient. Inspired by this, SLIMP combines the appearance and the metadata of individual skin lesions with patient-level metadata relating to their medical record and other clinically relevant information. By fully exploiting all available data modalities throughout the learning process, the proposed pre-training strategy improves performance compared to other pre-training strategies on downstream skin lesions classification tasks highlighting the learned representations quality.
Authors: Zeinab Nezami, Syed Danial Ali Shah, Maryam Hafeez, Karim Djemame, Syed Ali Raza Zaidi
Abstract: This paper envisions 6G as a self-evolving telecom ecosystem, where AI-driven intelligence enables dynamic adaptation beyond static connectivity. We explore the key enablers of autonomous communication systems, spanning reconfigurable infrastructure, adaptive middleware, and intelligent network functions, alongside multi-agent collaboration for distributed decision-making. We explore how these methodologies align with emerging industrial IoT frameworks, ensuring seamless integration within digital manufacturing processes. Our findings emphasize the potential for improved real-time decision-making, optimizing efficiency, and reducing latency in networked control systems. The discussion addresses ethical challenges, research directions, and standardization efforts, concluding with a technology stack roadmap to guide future developments. By leveraging state-of-the-art 6G network management techniques, this research contributes to the next generation of intelligent automation solutions, bridging the gap between theoretical advancements and real-world industrial applications.
Authors: Roksana Goworek, Harpal Karlcut, Muhammad Shezad, Nijaguna Darshana, Abhishek Mane, Syam Bondada, Raghav Sikka, Ulvi Mammadov, Rauf Allahverdiyev, Sriram Purighella, Paridhi Gupta, Muhinyia Ndegwa, Haim Dubossarsky
Abstract: This paper addresses the critical need for high-quality evaluation datasets in low-resource languages to advance cross-lingual transfer. While cross-lingual transfer offers a key strategy for leveraging multilingual pretraining to expand language technologies to understudied and typologically diverse languages, its effectiveness is dependent on quality and suitable benchmarks. We release new sense-annotated datasets of sentences containing polysemous words, spanning nine low-resource languages across diverse language families and scripts. To facilitate dataset creation, the paper presents a demonstrably beneficial semi-automatic annotation method. The utility of the datasets is demonstrated through Word-in-Context (WiC) formatted experiments that evaluate transfer on these low-resource languages. Results highlight the importance of targeted dataset creation and evaluation for effective polysemy disambiguation in low-resource settings and transfer studies. The released datasets and code aim to support further research into fair, robust, and truly multilingual NLP.
Authors: Arun Verma, Indrajit Saha, Makoto Yokoo, Bryan Kian Hsiang Low
Abstract: This paper considers a contextual bandit problem involving multiple agents, where a learner sequentially observes the contexts and the agent's reported arms, and then selects the arm that maximizes the system's overall reward. Existing work in contextual bandits assumes that agents truthfully report their arms, which is unrealistic in many real-life applications. For instance, consider an online platform with multiple sellers; some sellers may misrepresent product quality to gain an advantage, such as having the platform preferentially recommend their products to online users. To address this challenge, we propose an algorithm, COBRA, for contextual bandit problems involving strategic agents that disincentivize their strategic behavior without using any monetary incentives, while having incentive compatibility and a sub-linear regret guarantee. Our experimental results also validate the different performance aspects of our proposed algorithm.
Authors: Zexi Liu, Jingyi Chai, Xinyu Zhu, Shuo Tang, Rui Ye, Bo Zhang, Lei Bai, Siheng Chen
Abstract: The emergence of large language model (LLM)-based agents has significantly advanced the development of autonomous machine learning (ML) engineering. However, most existing approaches rely heavily on manual prompt engineering, failing to adapt and optimize based on diverse experimental experiences. Focusing on this, for the first time, we explore the paradigm of learning-based agentic ML, where an LLM agent learns through interactive experimentation on ML tasks using online reinforcement learning (RL). To realize this, we propose a novel agentic ML training framework with three key components: (1) exploration-enriched fine-tuning, which enables LLM agents to generate diverse actions for enhanced RL exploration; (2) step-wise RL, which enables training on a single action step, accelerating experience collection and improving training efficiency; (3) an agentic ML-specific reward module, which unifies varied ML feedback signals into consistent rewards for RL optimization. Leveraging this framework, we train ML-Agent, driven by a 7B-sized Qwen-2.5 LLM for autonomous ML. Remarkably, despite being trained on merely 9 ML tasks, our 7B-sized ML-Agent outperforms the 671B-sized DeepSeek-R1 agent. Furthermore, it achieves continuous performance improvements and demonstrates exceptional cross-task generalization capabilities.
Authors: Minrui Luo, Fuhang Kuang, Yu Wang, Zirui Liu, Tianxing He
Abstract: Parameter-Efficient Fine-Tuning (PEFT) methods, particularly Low-Rank Adaptation (LoRA), are indispensable for efficiently customizing Large Language Models (LLMs). However, vanilla LoRA suffers from slow convergence speed and knowledge forgetting problems. Recent studies have leveraged the power of designed LoRA initialization, to enhance the fine-tuning efficiency, or to preserve knowledge in the pre-trained LLM. However, none of these works can address the two cases at the same time. To this end, we introduce Subspace-Constrained LoRA (SC-LoRA), a novel LoRA initialization framework engineered to navigate the trade-off between efficient fine-tuning and knowledge preservation. We achieve this by constraining the output of trainable LoRA adapters in a low-rank subspace, where the context information of fine-tuning data is most preserved while the context information of preserved knowledge is least retained, in a balanced way. Such constraint enables the trainable weights to primarily focus on the main features of fine-tuning data while avoiding damaging the preserved knowledge features. We provide theoretical analysis on our method, and conduct extensive experiments including safety preservation and world knowledge preservation, on various downstream tasks. In our experiments, SC-LoRA succeeds in delivering superior fine-tuning performance while markedly diminishing knowledge forgetting, surpassing contemporary LoRA initialization methods.
Authors: Mohamad Chehade, Soumya Suvra Ghosal, Souradip Chakraborty, Avinash Reddy, Dinesh Manocha, Hao Zhu, Amrit Singh Bedi
Abstract: Aligning large language models with humans is challenging due to the inherently multifaceted nature of preference feedback. While existing approaches typically frame this as a multi-objective optimization problem, they often overlook how humans actually make decisions. Research on bounded rationality suggests that human decision making follows satisficing strategies-optimizing primary objectives while ensuring others meet acceptable thresholds. To bridge this gap and operationalize the notion of satisficing alignment, we propose SITAlign: an inference time framework that addresses the multifaceted nature of alignment by maximizing a primary objective while satisfying threshold-based constraints on secondary criteria. We provide theoretical insights by deriving sub-optimality bounds of our satisficing based inference alignment approach. We empirically validate SITAlign's performance through extensive experimentation on multiple benchmarks. For instance, on the PKU-SafeRLHF dataset with the primary objective of maximizing helpfulness while ensuring a threshold on harmlessness, SITAlign outperforms the state-of-the-art multi objective decoding strategy by a margin of 22.3% in terms of GPT-4 win-tie rate for helpfulness reward while adhering to the threshold on harmlessness.
Authors: Truong (Jack), Luu, Binny M. Samuel
Abstract: In recent years, the rapid advancement and democratization of generative AI models have sparked significant debate over safety, ethical risks, and dual-use concerns, particularly in the context of cybersecurity. While anecdotally known, this paper provides empirical evidence regarding generative AI's association with malicious internet-related activities and cybercrime by examining the phenomenon through psychological frameworks of technological amplification and affordance theory. Using a quasi-experimental design with interrupted time series analysis, we analyze two datasets, one general and one cryptocurrency-focused, to empirically assess generative AI's role in cybercrime. The findings contribute to ongoing discussions about AI governance by balancing control and fostering innovation, underscoring the need for strategies to guide policymakers, inform AI developers and cybersecurity professionals, and educate the public to maximize AI's benefits while mitigating its risks.
Authors: Ali Behrouz, Zeman Li, Praneeth Kacham, Majid Daliri, Yuan Deng, Peilin Zhong, Meisam Razaviyayn, Vahab Mirrokni
Abstract: Transformers have been established as the most popular backbones in sequence modeling, mainly due to their effectiveness in in-context retrieval tasks and the ability to learn at scale. Their quadratic memory and time complexity, however, bound their applicability in longer sequences and so has motivated researchers to explore effective alternative architectures such as modern recurrent neural networks (a.k.a long-term recurrent memory module). Despite their recent success in diverse downstream tasks, they struggle in tasks that requires long context understanding and extrapolation to longer sequences. We observe that these shortcomings come from three disjoint aspects in their design: (1) limited memory capacity that is bounded by the architecture of memory and feature mapping of the input; (2) online nature of update, i.e., optimizing the memory only with respect to the last input; and (3) less expressive management of their fixed-size memory. To enhance all these three aspects, we present ATLAS, a long-term memory module with high capacity that learns to memorize the context by optimizing the memory based on the current and past tokens, overcoming the online nature of long-term memory models. Building on this insight, we present a new family of Transformer-like architectures, called DeepTransformers, that are strict generalizations of the original Transformer architecture. Our experimental results on language modeling, common-sense reasoning, recall-intensive, and long-context understanding tasks show that ATLAS surpasses the performance of Transformers and recent linear recurrent models. ATLAS further improves the long context performance of Titans, achieving +80\% accuracy in 10M context length of BABILong benchmark.
Authors: Yufan Deng, Xun Guo, Yuanyang Yin, Jacob Zhiyuan Fang, Yiding Yang, Yizhi Wang, Shenghai Yuan, Angtian Wang, Bo Liu, Haibin Huang, Chongyang Ma
Abstract: Video generation has made substantial strides with the emergence of deep generative models, especially diffusion-based approaches. However, video generation based on multiple reference subjects still faces significant challenges in maintaining multi-subject consistency and ensuring high generation quality. In this paper, we propose MAGREF, a unified framework for any-reference video generation that introduces masked guidance to enable coherent multi-subject video synthesis conditioned on diverse reference images and a textual prompt. Specifically, we propose (1) a region-aware dynamic masking mechanism that enables a single model to flexibly handle various subject inference, including humans, objects, and backgrounds, without architectural changes, and (2) a pixel-wise channel concatenation mechanism that operates on the channel dimension to better preserve appearance features. Our model delivers state-of-the-art video generation quality, generalizing from single-subject training to complex multi-subject scenarios with coherent synthesis and precise control over individual subjects, outperforming existing open-source and commercial baselines. To facilitate evaluation, we also introduce a comprehensive multi-subject video benchmark. Extensive experiments demonstrate the effectiveness of our approach, paving the way for scalable, controllable, and high-fidelity multi-subject video synthesis. Code and model can be found at: https://github.com/MAGREF-Video/MAGREF
Authors: Qiang Wang, Xiang Song, Yuhang He, Jizhou Han, Chenhao Ding, Xinyuan Gao, Yihong Gong
Abstract: Deep neural networks (DNNs) often underperform in real-world, dynamic settings where data distributions change over time. Domain Incremental Learning (DIL) offers a solution by enabling continual model adaptation, with Parameter-Isolation DIL (PIDIL) emerging as a promising paradigm to reduce knowledge conflicts. However, existing PIDIL methods struggle with parameter selection accuracy, especially as the number of domains and corresponding classes grows. To address this, we propose SOYO, a lightweight framework that improves domain selection in PIDIL. SOYO introduces a Gaussian Mixture Compressor (GMC) and Domain Feature Resampler (DFR) to store and balance prior domain data efficiently, while a Multi-level Domain Feature Fusion Network (MDFN) enhances domain feature extraction. Our framework supports multiple Parameter-Efficient Fine-Tuning (PEFT) methods and is validated across tasks such as image classification, object detection, and speech enhancement. Experimental results on six benchmarks demonstrate SOYO's consistent superiority over existing baselines, showcasing its robustness and adaptability in complex, evolving environments. The codes will be released in https://github.com/qwangcv/SOYO.
Authors: Hao Dong, Moru Liu, Jian Liang, Eleni Chatzi, Olga Fink
Abstract: Vision-Language Models (VLMs) have demonstrated strong capabilities in aligning visual and textual modalities, enabling a wide range of applications in multimodal understanding and generation. While they excel in zero-shot and transfer learning scenarios, VLMs remain susceptible to misclassification, often yielding confident yet incorrect predictions. This limitation poses a significant risk in safety-critical domains, where erroneous predictions can lead to severe consequences. In this work, we introduce TrustVLM, a training-free framework designed to address the critical challenge of estimating when VLM's predictions can be trusted. Motivated by the observed modality gap in VLMs and the insight that certain concepts are more distinctly represented in the image embedding space, we propose a novel confidence-scoring function that leverages this space to improve misclassification detection. We rigorously evaluate our approach across 17 diverse datasets, employing 4 architectures and 2 VLMs, and demonstrate state-of-the-art performance, with improvements of up to 51.87% in AURC, 9.14% in AUROC, and 32.42% in FPR95 compared to existing baselines. By improving the reliability of the model without requiring retraining, TrustVLM paves the way for safer deployment of VLMs in real-world applications. The code will be available at https://github.com/EPFL-IMOS/TrustVLM.
Authors: Diankun Wu, Fangfu Liu, Yi-Hsin Hung, Yueqi Duan
Abstract: Recent advancements in Multimodal Large Language Models (MLLMs) have significantly enhanced performance on 2D visual tasks. However, improving their spatial intelligence remains a challenge. Existing 3D MLLMs always rely on additional 3D or 2.5D data to incorporate spatial awareness, restricting their utility in scenarios with only 2D inputs, such as images or videos. In this paper, we present Spatial-MLLM, a novel framework for visual-based spatial reasoning from purely 2D observations. Unlike conventional video MLLMs which rely on CLIP-based visual encoders optimized for semantic understanding, our key insight is to unleash the strong structure prior from the feed-forward visual geometry foundation model. Specifically, we propose a dual-encoder architecture: a pretrained 2D visual encoder to extract semantic features, and a spatial encoder-initialized from the backbone of the visual geometry model-to extract 3D structure features. A connector then integrates both features into unified visual tokens for enhanced spatial understanding. Furthermore, we propose a space-aware frame sampling strategy at inference time, which selects the spatially informative frames of a video sequence, ensuring that even under limited token length, the model focuses on frames critical for spatial reasoning. Beyond architecture improvements, we construct the Spatial-MLLM-120k dataset and train the model on it using supervised fine-tuning and GRPO. Extensive experiments on various real-world datasets demonstrate that our spatial-MLLM achieves state-of-the-art performance in a wide range of visual-based spatial understanding and reasoning tasks. Project page: https://diankun-wu.github.io/Spatial-MLLM/.
Authors: Declan Kutscher, David M. Chan, Yutong Bai, Trevor Darrell, Ritwik Gupta
Abstract: Sequence models such as transformers require inputs to be represented as one-dimensional sequences. In vision, this typically involves flattening images using a fixed row-major (raster-scan) order. While full self-attention is permutation-equivariant, modern long-sequence transformers increasingly rely on architectural approximations that break this invariance and introduce sensitivity to patch ordering. We show that patch order significantly affects model performance in such settings, with simple alternatives like column-major or Hilbert curves yielding notable accuracy shifts. Motivated by this, we propose REOrder, a two-stage framework for discovering task-optimal patch orderings. First, we derive an information-theoretic prior by evaluating the compressibility of various patch sequences. Then, we learn a policy over permutations by optimizing a Plackett-Luce policy using REINFORCE. This approach enables efficient learning in a combinatorial permutation space. REOrder improves top-1 accuracy over row-major ordering on ImageNet-1K by up to 3.01% and Functional Map of the World by 13.35%.
Authors: Ziyin Zhang, Jiahao Xu, Zhiwei He, Tian Liang, Qiuzhi Liu, Yansi Li, Linfeng Song, Zhengwen Liang, Zhuosheng Zhang, Rui Wang, Zhaopeng Tu, Haitao Mi, Dong Yu
Abstract: Theorem proving serves as a major testbed for evaluating complex reasoning abilities in large language models (LLMs). However, traditional automated theorem proving (ATP) approaches rely heavily on formal proof systems that poorly align with LLMs' strength derived from informal, natural language knowledge acquired during pre-training. In this work, we propose DeepTheorem, a comprehensive informal theorem-proving framework exploiting natural language to enhance LLM mathematical reasoning. DeepTheorem includes a large-scale benchmark dataset consisting of 121K high-quality IMO-level informal theorems and proofs spanning diverse mathematical domains, rigorously annotated for correctness, difficulty, and topic categories, accompanied by systematically constructed verifiable theorem variants. We devise a novel reinforcement learning strategy (RL-Zero) explicitly tailored to informal theorem proving, leveraging the verified theorem variants to incentivize robust mathematical inference. Additionally, we propose comprehensive outcome and process evaluation metrics examining proof correctness and the quality of reasoning steps. Extensive experimental analyses demonstrate DeepTheorem significantly improves LLM theorem-proving performance compared to existing datasets and supervised fine-tuning protocols, achieving state-of-the-art accuracy and reasoning quality. Our findings highlight DeepTheorem's potential to fundamentally advance automated informal theorem proving and mathematical exploration.
Authors: Heekyung Lee, Jiaxin Ge, Tsung-Han Wu, Minwoo Kang, Trevor Darrell, David M. Chan
Abstract: Rebus puzzles, visual riddles that encode language through imagery, spatial arrangement, and symbolic substitution, pose a unique challenge to current vision-language models (VLMs). Unlike traditional image captioning or question answering tasks, rebus solving requires multi-modal abstraction, symbolic reasoning, and a grasp of cultural, phonetic and linguistic puns. In this paper, we investigate the capacity of contemporary VLMs to interpret and solve rebus puzzles by constructing a hand-generated and annotated benchmark of diverse English-language rebus puzzles, ranging from simple pictographic substitutions to spatially-dependent cues ("head" over "heels"). We analyze how different VLMs perform, and our findings reveal that while VLMs exhibit some surprising capabilities in decoding simple visual clues, they struggle significantly with tasks requiring abstract reasoning, lateral thinking, and understanding visual metaphors.
Authors: Yunjae Won, Hyunji Lee, Hyeonbin Hwang, Minjoon Seo
Abstract: Direct Preference Optimization (DPO) has become a standard technique for aligning language models with human preferences in a supervised manner. Despite its empirical success, the theoretical justification behind its log-ratio reward parameterization remains incomplete. In this work, we address this gap by utilizing the Differential Information Distribution (DID): a distribution over token sequences that captures the information gained during policy updates. First, we show that when preference labels encode the differential information required to transform a reference policy into a target policy, the log-ratio reward in DPO emerges as the uniquely optimal form for learning the target policy via preference optimization. This result naturally yields a closed-form expression for the optimal sampling distribution over rejected responses. Second, we find that the condition for preferences to encode differential information is fundamentally linked to an implicit assumption regarding log-margin ordered policies-an inductive bias widely used in preference optimization yet previously unrecognized. Finally, by analyzing the entropy of the DID, we characterize how learning low-entropy differential information reinforces the policy distribution, while high-entropy differential information induces a smoothing effect, which explains the log-likelihood displacement phenomenon. We validate our theoretical findings in synthetic experiments and extend them to real-world instruction-following datasets. Our results suggest that learning high-entropy differential information is crucial for general instruction-following, while learning low-entropy differential information benefits knowledge-intensive question answering. Overall, our work presents a unifying perspective on the DPO objective, the structure of preference data, and resulting policy behaviors through the lens of differential information.
Authors: Wentao Zhang, Woojeong Kim, Yuntian Deng
Abstract: Conversational agents powered by large language models (LLMs) are rapidly becoming integral to our daily interactions, generating unprecedented amounts of conversational data. Such datasets offer a powerful lens into societal interests, trending topics, and collective concerns. Yet, existing approaches typically treat these interactions as independent and miss critical insights that could emerge from aggregating and reasoning across large-scale conversation logs. In this paper, we introduce Aggregative Question Answering, a novel task requiring models to reason explicitly over thousands of user-chatbot interactions to answer aggregative queries, such as identifying emerging concerns among specific demographics. To enable research in this direction, we construct a benchmark, WildChat-AQA, comprising 6,027 aggregative questions derived from 182,330 real-world chatbot conversations. Experiments show that existing methods either struggle to reason effectively or incur prohibitive computational costs, underscoring the need for new approaches capable of extracting collective insights from large-scale conversational data.
Authors: Zelai Xu, Chao Yu, Fei Fang, Yu Wang, Yi Wu
Abstract: Agents built with large language models (LLMs) have shown great potential across a wide range of domains. However, in complex decision-making tasks, pure LLM-based agents tend to exhibit intrinsic bias in their choice of actions, which is inherited from the model's training data and results in suboptimal performance. To develop strategic language agents, i.e., agents that generate flexible language actions and possess strong decision-making abilities, we propose a novel framework that powers LLM-based agents with reinforcement learning (RL). We consider Werewolf, a popular social deduction game, as a challenging testbed that emphasizes versatile communication and strategic gameplay. To mitigate the intrinsic bias in language actions, our agents use an LLM to perform deductive reasoning and generate a diverse set of action candidates. Then an RL policy trained to optimize the decision-making ability chooses an action from the candidates to play in the game. Extensive experiments show that our agents overcome the intrinsic bias and outperform existing LLM-based agents in the Werewolf game. We also conduct human-agent experiments and find that our agents achieve human-level performance and demonstrate strong strategic play.
Authors: Hendrik Buschmeier, Heike M. Buhl, Friederike Kern, Angela Grimminger, Helen Beierling, Josephine Fisher, Andr\'e Gro{\ss}, Ilona Horwath, Nils Klowait, Stefan Lazarov, Michael Lenke, Vivien Lohmer, Katharina Rohlfing, Ingrid Scharlau, Amit Singh, Lutz Terfloth, Anna-Lisa Vollmer, Yu Wang, Annedore Wilmes, Britta Wrede
Abstract: Explainability has become an important topic in computer science and artificial intelligence, leading to a subfield called Explainable Artificial Intelligence (XAI). The goal of providing or seeking explanations is to achieve (better) 'understanding' on the part of the explainee. However, what it means to 'understand' is still not clearly defined, and the concept itself is rarely the subject of scientific investigation. This conceptual article aims to present a model of forms of understanding for XAI-explanations and beyond. From an interdisciplinary perspective bringing together computer science, linguistics, sociology, philosophy and psychology, a definition of understanding and its forms, assessment, and dynamics during the process of giving everyday explanations are explored. Two types of understanding are considered as possible outcomes of explanations, namely enabledness, 'knowing how' to do or decide something, and comprehension, 'knowing that' -- both in different degrees (from shallow to deep). Explanations regularly start with shallow understanding in a specific domain and can lead to deep comprehension and enabledness of the explanandum, which we see as a prerequisite for human users to gain agency. In this process, the increase of comprehension and enabledness are highly interdependent. Against the background of this systematization, special challenges of understanding in XAI are discussed.
Authors: Shengjie Ma, Qi Chu, Jiaxin Mao, Xuhui Jiang, Haozhe Duan, Chong Chen
Abstract: Determining which legal cases are relevant to a given query involves navigating lengthy texts and applying nuanced legal reasoning. Traditionally, this task has demanded significant time and domain expertise to identify key Legal Facts and reach sound juridical conclusions. In addition, existing data with legal case similarities often lack interpretability, making it difficult to understand the rationale behind relevance judgments. With the growing capabilities of large language models (LLMs), researchers have begun investigating their potential in this domain. Nonetheless, the method of employing a general large language model for reliable relevance judgments in legal case retrieval remains largely unexplored. To address this gap in research, we propose a novel few-shot approach where LLMs assist in generating expert-aligned interpretable relevance judgments. The proposed approach decomposes the judgment process into several stages, mimicking the workflow of human annotators and allowing for the flexible incorporation of expert reasoning to improve the accuracy of relevance judgments. Importantly, it also ensures interpretable data labeling, providing transparency and clarity in the relevance assessment process. Through a comparison of relevance judgments made by LLMs and human experts, we empirically demonstrate that the proposed approach can yield reliable and valid relevance assessments. Furthermore, we demonstrate that with minimal expert supervision, our approach enables a large language model to acquire case analysis expertise and subsequently transfers this ability to a smaller model via annotation-based knowledge distillation.
Authors: Yongqiang Yao, Jingru Tan, Feizhao Zhang, Jiahao Hu, Yazhe Niu, Xin Jin, Bo Li, Pengfei Liu, Ruihao Gong, Dahua Lin, Ningyi Xu
Abstract: Vision-language instruction-tuning models have recently achieved significant performance improvements. In this work, we discover that large-scale 3D parallel training on those models leads to an imbalanced computation load across different devices. The vision and language parts are inherently heterogeneous: their data distribution and model architecture differ significantly, which affects distributed training efficiency. To address this issue, we rebalance the computational load from data, model, and memory perspectives, achieving more balanced computation across devices. Specifically, for the data, instances are grouped into new balanced mini-batches within and across devices. A search-based method is employed for the model to achieve a more balanced partitioning. For memory optimization, we adaptively adjust the re-computation strategy for each partition to utilize the available memory fully. These three perspectives are not independent but are closely connected, forming an omniverse balanced training framework. Extensive experiments are conducted to validate the effectiveness of our method. Compared with the open-source training code of InternVL-Chat, training time is reduced greatly, achieving about 1.8$\times$ speed-up. Our method's efficacy and generalizability are further validated across various models and datasets. Codes will be released at https://github.com/ModelTC/OmniBal.
Authors: Jen-tse Huang, Jiaxu Zhou, Tailin Jin, Xuhui Zhou, Zixi Chen, Wenxuan Wang, Youliang Yuan, Michael R. Lyu, Maarten Sap
Abstract: Large language model-based multi-agent systems have shown great abilities across various tasks due to the collaboration of expert agents, each focusing on a specific domain. However, the impact of clumsy or even malicious agents--those who frequently make errors in their tasks--on the overall performance of the system remains underexplored. This paper investigates: (1) What is the resilience of various system structures (e.g., A$\rightarrow$B$\rightarrow$C, A$\leftrightarrow$B$\leftrightarrow$C) under faulty agents, on different downstream tasks? (2) How can we increase system resilience to defend against these agents? To simulate faulty agents, we propose two approaches--AutoTransform and AutoInject--which introduce mistakes into the agents' responses. Experiments on four downstream tasks using six systems show that the "hierarchical" structure, i.e., A$\rightarrow$(B$\leftrightarrow$C), exhibits superior resilience with the lowest performance drop of 5.5%, compared to 10.5% and 23.7% of other two structures. To further improve resilience, we introduce (1) Challenger, that introduces a mechanism for each agent to challenge others' outputs, and (2) Inspector, an additional agent to review and correct messages, recovering up to 96.4% errors made by faulty agents. Our code and data are available at https://github.com/CUHK-ARISE/MAS-Resilience.
Authors: Anton Korikov, Scott Sanner, Yashar Deldjoo, Zhankui He, Julian McAuley, Arnau Ramisa, Rene Vidal, Mahesh Sathiamoorthy, Atoosa Kasrizadeh, Silvia Milano, Francesco Ricci
Abstract: While previous chapters focused on recommendation systems (RSs) based on standardized, non-verbal user feedback such as purchases, views, and clicks -- the advent of LLMs has unlocked the use of natural language (NL) interactions for recommendation. This chapter discusses how LLMs' abilities for general NL reasoning present novel opportunities to build highly personalized RSs -- which can effectively connect nuanced and diverse user preferences to items, potentially via interactive dialogues. To begin this discussion, we first present a taxonomy of the key data sources for language-driven recommendation, covering item descriptions, user-system interactions, and user profiles. We then proceed to fundamental techniques for LLM recommendation, reviewing the use of encoder-only and autoregressive LLM recommendation in both tuned and untuned settings. Afterwards, we move to multi-module recommendation architectures in which LLMs interact with components such as retrievers and RSs in multi-stage pipelines. This brings us to architectures for conversational recommender systems (CRSs), in which LLMs facilitate multi-turn dialogues where each turn presents an opportunity not only to make recommendations, but also to engage with the user in interactive preference elicitation, critiquing, and question-answering.
Authors: Agnieszka Mensfelt, Kostas Stathis, Vince Trencsenyi
Abstract: Large language models (LLMs) are increasingly explored as general-purpose reasoners, particularly in agentic contexts. However, their outputs remain prone to mathematical and logical errors. This is especially challenging in open-ended tasks, where unstructured outputs lack explicit ground truth and may contain subtle inconsistencies. To address this issue, we propose Logic-Enhanced Language Model Agents (LELMA), a framework that integrates LLMs with formal logic to enable validation and refinement of natural language reasoning. LELMA comprises three components: an LLM-Reasoner, an LLM-Translator, and a Solver, and employs autoformalization to translate reasoning into logic representations, which are then used to assess logical validity. Using game-theoretic scenarios such as the Prisoner's Dilemma as testbeds, we highlight the limitations of both less capable (Gemini 1.0 Pro) and advanced (GPT-4o) models in generating logically sound reasoning. LELMA achieves high accuracy in error detection and improves reasoning correctness via self-refinement, particularly in GPT-4o. The study also highlights challenges in autoformalization accuracy and in evaluation of inherently ambiguous open-ended reasoning tasks.
Authors: Huimu Yu, Xing Wu, Haotian Xu, Debing Zhang, Songlin Hu
Abstract: Large language models (LLMs) have made significant progress in natural language understanding and generation, driven by scalable pretraining and advanced finetuning. However, enhancing reasoning abilities in LLMs, particularly via reinforcement learning from human feedback (RLHF), remains challenging due to the scarcity of high-quality preference data, which is labor-intensive to annotate and crucial for reward model (RM) finetuning. To alleviate this issue, we introduce CodePMP, a scalable preference model pretraining (PMP) pipeline that utilizes a large corpus of synthesized code-preference pairs from publicly available high-quality source code. CodePMP improves RM finetuning efficiency by pretraining preference models on large-scale synthesized code-preference pairs. We evaluate CodePMP on mathematical reasoning tasks (GSM8K, MATH) and logical reasoning tasks (ReClor, LogiQA2.0), consistently showing significant improvements in reasoning performance of LLMs and highlighting the importance of scalable preference model pretraining for efficient reward modeling.
Authors: Jiashu He, Mingyu Derek Ma, Jinxuan Fan, Dan Roth, Wei Wang, Alejandro Ribeiro
Abstract: Existing approaches based on context prompting or reinforcement learning (RL) to improve the reasoning capacities of large language models (LLMs) depend on the LLMs' internal knowledge to produce reliable Chain-Of-Thought (CoT). However, no matter the size of LLMs, certain problems cannot be resolved in a single forward pass. Meanwhile, agent-based reasoning systems require access to a comprehensive nonparametric knowledge base, which is often costly or not feasible for use in scientific and niche domains. We present Graph Inspired Veracity Extrapolation (GIVE), a novel reasoning method that merges parametric and non-parametric memories to improve accurate reasoning with minimal external input. GIVE guides the LLM agent to select the most pertinent expert data (observe), engage in query-specific divergent thinking (reflect), and then synthesize this information to produce the final output (speak). Extensive experiments demonstrated the following benefits of our framework: (1) GIVE boosts the performance of LLMs across various sizes. (2) In some scenarios, GIVE allows smaller LLMs to surpass larger, more sophisticated ones in scientific tasks (GPT3.5T + GIVE > GPT4). (3) GIVE is effective on scientific and open-domain assessments. (4) GIVE is a training-free method that enables LLMs to tackle new problems that extend beyond their training data (up to 43.5% -> 88.2%} accuracy improvement). (5) GIVE allows LLM agents to reason using both restricted (very small) and noisy (very large) knowledge sources, accommodating knowledge graphs (KG) ranging from 135 to more than 840k nodes. (6) The reasoning process involved in GIVE is fully interpretable.
Authors: Linger Deng, Linghao Zhu, Yuliang Liu, Yu Wang, Qunyi Xie, Jingjing Wu, Gang Zhang, Yingying Zhu, Xiang Bai
Abstract: Large Multimodal Models (LMMs) face limitations in geometric reasoning due to insufficient Chain of Thought (CoT) image-text training data. While existing approaches leverage template-based or LLM-assisted methods for geometric CoT data creation, they often face challenges in achieving both diversity and precision. To bridge this gap, we introduce a two-stage Theorem-Validated Reverse Chain-of-Thought Reasoning Synthesis (TR-CoT) framework. The first stage, TR-Engine, synthesizes theorem-grounded geometric diagrams with structured descriptions and properties. The second stage, TR-Reasoner, employs reverse reasoning to iteratively refine question-answer pairs by cross-validating geometric properties and description fragments. Our approach expands theorem-type coverage, corrects long-standing misunderstandings, and enhances geometric reasoning. Fine-grained CoT improves theorem understanding and increases logical consistency by 24.5%. Our best models surpass the baselines in MathVista and GeoQA by 10.1% and 4.7%, outperforming advanced closed-source models like GPT-4o.
Authors: Agnieszka Mensfelt, Kostas Stathis, Vince Trencsenyi
Abstract: Multi-agent simulations are versatile tools for exploring interactions among natural and artificial agents, but their development typically demands domain expertise and manual effort. This work introduces the Generative Agents for Multi-Agent Autoformalization (GAMA) framework, which automates the formalization of interaction scenarios in simulations using agents augmented with large language models (LLMs). To demonstrate the application of GAMA, we use natural language descriptions of game-theoretic scenarios representing social interactions, and we autoformalize them into executable logic programs defining game rules, with syntactic correctness enforced through a solver-based validation. To ensure runtime validity, an iterative, tournament-based procedure tests the generated rules and strategies, followed by exact semantic validation when ground truth outcomes are available. In experiments with 110 natural language descriptions across five 2x2 simultaneous-move games, GAMA achieves 100% syntactic and 76.5% semantic correctness with Claude 3.5 Sonnet, and 99.82% syntactic and 77% semantic correctness with GPT-4o. The framework also shows high semantic accuracy in autoformalizing agents' strategies.
Authors: Shaofei Cai, Zhancun Mu, Kaichen He, Bowei Zhang, Xinyue Zheng, Anji Liu, Yitao Liang
Abstract: Minecraft's complexity and diversity as an open world make it a perfect environment to test if agents can learn, adapt, and tackle a variety of unscripted tasks. However, the development and validation of novel agents in this setting continue to face significant engineering challenges. This paper presents MineStudio, an open-source software package designed to streamline the development of autonomous agents in Minecraft. MineStudio represents the first comprehensive integration of seven critical engineering components: simulator, data, model, offline pre-training, online fine-tuning, inference, and benchmark, thereby allowing users to concentrate their efforts on algorithm innovation. We provide a user-friendly API design accompanied by comprehensive documentation and tutorials. Our project is released at https://github.com/CraftJarvis/MineStudio.
Authors: Nolan Koblischke, Hyunseok Jang, Kristen Menou, Mohamad Ali-Dib
Abstract: Modern science emerged from reasoning over repeatedly-observed planetary motions. We present Gravity-Bench-v1, an environment-based benchmark that challenges AI agents on tasks that parallel this historical development. Gravity-Bench-v1 evaluates agents on the discovery of physics concealed within a dynamic environment, using rigorous gravitational dynamics simulations. Gravity-Bench includes out-of-distribution cases, i.e. with physics that deviates from the real world, to evaluate true scientific generalization capabilities. Agents must plan to collect data within an experimental budget and must perform a dynamic form of data analysis and reasoning to solve tasks efficiently. Our benchmark admits an open-ended space of solutions. Reference solutions for each task are provided to calibrate AI performance against human expertise. Technically at an upper-undergraduate level, our benchmark proves challenging to baseline AI agents. Gravity-Bench-v1 and planned extensions should help map out AI progress towards scientific discovery capabilities.
Authors: Anirudh Chari, Suraj Reddy, Aditya Tiwari, Richard Lian, Brian Zhou
Abstract: While large language models (LLMs) have shown promising capabilities as zero-shot planners for embodied agents, their inability to learn from experience and build persistent mental models limits their robustness in complex open-world environments like Minecraft. We introduce MINDSTORES, an experience-augmented planning framework that enables embodied agents to build and leverage mental models through natural interaction with their environment. Drawing inspiration from how humans construct and refine cognitive mental models, our approach extends existing zero-shot LLM planning by maintaining a database of past experiences that informs future planning iterations. The key innovation is representing accumulated experiences as natural language embeddings of (state, task, plan, outcome) tuples, which can then be efficiently retrieved and reasoned over by an LLM planner to generate insights and guide plan refinement for novel states and tasks. Through extensive experiments in the MineDojo environment, a simulation environment for agents in Minecraft that provides low-level controls for Minecraft, we find that MINDSTORES learns and applies its knowledge significantly better than existing memory-based LLM planners while maintaining the flexibility and generalization benefits of zero-shot approaches, representing an important step toward more capable embodied AI systems that can learn continuously through natural experience.
Authors: Ziyang Guo, Yifan Wu, Jason Hartline, Jessica Hullman
Abstract: Multiple agents -- including humans and AI models -- are increasingly combined to make decisions with the expectation of achieving complementary performance, where the decisions they make together outperform those made individually. However, knowing how to improve the performance of collaborating agents is often difficult without knowing more about what particular information and strategies each agent employs. With a focus on human-AI pairings, we contribute a decision-theoretic framework for characterizing the value of information -- and consequently, opportunities for agents to better exploit available information -- in AI-assisted decision workflows. We present a novel explanation technique (ILIV-SHAP) that adapts SHAP explanations to highlight human-complementing information. We validate the effectiveness of the framework and ILIV-SHAP through a study of human-AI decision-making. We show that our measure of complementary information can be used to identify which AI model will best complement human decisions. We also find that presenting ILIV-SHAP with AI predictions leads to reliably greater reductions in error over non-AI assisted decisions more than vanilla SHAP.
Authors: Xiandong Zou, Wanyu Lin, Yuchen Li, Pan Zhou
Abstract: Aligning Large Language Model (LLM) responses with human preferences is vital for building safe and controllable AI systems. While preference optimization methods based on Plackett-Luce (PL) and Bradley-Terry (BT) models have shown promise, they face challenges such as poor handling of harmful content, inefficient use of dispreferred responses, and, specifically for PL, high computational costs. To address these issues, we propose Hard Preference Sampling (HPS), a novel framework for robust and efficient human preference alignment. HPS introduces a training loss that prioritizes the most preferred response while rejecting all dispreferred and harmful ones. It emphasizes "hard" dispreferred responses -- those closely resembling preferred ones -- to enhance the model's rejection capabilities. By leveraging a single-sample Monte Carlo sampling strategy, HPS reduces computational overhead while maintaining alignment quality. Theoretically, HPS improves sample efficiency over existing PL methods and maximizes the reward margin between preferred and dispreferred responses, ensuring clearer distinctions. Experiments on HH-RLHF and PKU-Safety datasets validate HPS's effectiveness, achieving comparable BLEU and reward scores while greatly improving reward margins and thus reducing harmful content generation.
Authors: Michal Bravansky, Vaclav Kubon, Suhas Hariharan, Robert Kirk
Abstract: Interpreting data is central to modern research. Large language models (LLMs) show promise in providing such natural language interpretations of data, yet simple feature extraction methods such as prompting often fail to produce accurate and versatile descriptions for diverse datasets and lack control over granularity and scale. To address these limitations, we propose a domain-agnostic method for dataset featurization that provides precise control over the number of features extracted while maintaining compact and descriptive representations comparable to human labeling. Our method optimizes the selection of informative binary features by evaluating the ability of an LLM to reconstruct the original data using those features. We demonstrate its effectiveness in dataset modeling tasks and through two case studies: (1) Constructing a feature representation of jailbreak tactics that compactly captures both the effectiveness and diversity of a larger set of human-crafted attacks; and (2) automating the discovery of features that align with human preferences, achieving accuracy and robustness comparable to human-crafted features. Moreover, we show that the pipeline scales effectively, improving as additional features are sampled, making it suitable for large and diverse datasets.
Authors: Xinjie Zhao, Fan Gao, Xingyu Song, Yingjian Chen, Rui Yang, Yanran Fu, Yuyang Wang, Yusuke Iwasawa, Yutaka Matsuo, Irene Li
Abstract: Recent advances in large language models (LLMs) have significantly improved multi-hop question answering (QA) through direct Chain-of-Thought (CoT) reasoning. However, the irreversible nature of CoT leads to error accumulation, making it challenging to correct mistakes in multi-hop reasoning. This paper introduces ReAgent: a Reversible multi-Agent collaborative framework augmented with explicit backtracking mechanisms, enabling reversible multi-hop reasoning. By incorporating text-based retrieval, information aggregation and validation, our system can detect and correct errors mid-reasoning, leading to more robust and interpretable QA outcomes. The framework and experiments serve as a foundation for future work on error-tolerant QA systems. Empirical evaluations across three benchmarks indicate ReAgent's efficacy, yielding average about 6\% improvements against baseline models.
Authors: Ali Baheri, Cecilia O. Alm
Abstract: We present a hierarchical neuro-symbolic control framework that tightly couples a classical symbolic planner with a transformer-based policy to address long-horizon decision-making under uncertainty. At the high level, the planner assembles an interpretable sequence of operators that guarantees logical coherence with task constraints, while at the low level each operator is rendered as a sub-goal token that conditions a decision transformer to generate fine-grained actions directly from raw observations. This bidirectional interface preserves the combinatorial efficiency and explainability of symbolic reasoning without sacrificing the adaptability of deep sequence models, and it permits a principled analysis that tracks how approximation errors from both planning and execution accumulate across the hierarchy. Empirical studies in stochastic grid-world domains demonstrate that the proposed method consistently surpasses purely symbolic, purely neural and existing hierarchical baselines in both success and efficiency, highlighting its robustness for sequential tasks.
Authors: Shilin Zhang, Zican Hu, Wenhao Wu, Xinyi Xie, Jianxiang Tang, Chunlin Chen, Daoyi Dong, Yu Cheng, Zhenhong Sun, Zhi Wang
Abstract: Offline meta-RL usually tackles generalization by inferring task beliefs from high-quality samples or warmup explorations. The restricted form limits their generality and usability since these supervision signals are expensive and even infeasible to acquire in advance for unseen tasks. Learning directly from the raw text about decision tasks is a promising alternative to leverage a much broader source of supervision. In the paper, we propose \textbf{T}ext-to-\textbf{D}ecision \textbf{A}gent (\textbf{T2DA}), a simple and scalable framework that supervises offline meta-RL with natural language. We first introduce a generalized world model to encode multi-task decision data into a dynamics-aware embedding space. Then, inspired by CLIP, we predict which textual description goes with which decision embedding, effectively bridging their semantic gap via contrastive language-decision pre-training and aligning the text embeddings to comprehend the environment dynamics. After training the text-conditioned generalist policy, the agent can directly realize zero-shot text-to-decision generation in response to language instructions. Comprehensive experiments on MuJoCo and Meta-World benchmarks show that T2DA facilitates high-capacity zero-shot generalization and outperforms various types of baselines.
Authors: D. Sculley, Will Cukierski, Phil Culliton, Sohier Dane, Maggie Demkin, Ryan Holbrook, Addison Howard, Paul Mooney, Walter Reade, Megan Risdal, Nate Keating
Abstract: In this position paper, we observe that empirical evaluation in Generative AI is at a crisis point since traditional ML evaluation and benchmarking strategies are insufficient to meet the needs of evaluating modern GenAI models and systems. There are many reasons for this, including the fact that these models typically have nearly unbounded input and output spaces, typically do not have a well defined ground truth target, and typically exhibit strong feedback loops and prediction dependence based on context of previous model outputs. On top of these critical issues, we argue that the problems of leakage and contamination are in fact the most important and difficult issues to address for GenAI evaluations. Interestingly, the field of AI Competitions has developed effective measures and practices to combat leakage for the purpose of counteracting cheating by bad actors within a competition setting. This makes AI Competitions an especially valuable (but underutilized) resource. Now is time for the field to view AI Competitions as the gold standard for empirical rigor in GenAI evaluation, and to harness and harvest their results with according value.
Authors: Runquan Gui, Zhihai Wang, Jie Wang, Chi Ma, Huiling Zhen, Mingxuan Yuan, Jianye Hao, Defu Lian, Enhong Chen, Feng Wu
Abstract: Recent advancements have significantly enhanced the performance of large language models (LLMs) in tackling complex reasoning tasks, achieving notable success in domains like mathematical and logical reasoning. However, these methods encounter challenges with complex planning tasks, primarily due to extended reasoning steps, diverse constraints, and the challenge of handling multiple distinct sub-tasks. To address these challenges, we propose HyperTree Planning (HTP), a novel reasoning paradigm that constructs hypertree-structured planning outlines for effective planning. The hypertree structure enables LLMs to engage in hierarchical thinking by flexibly employing the divide-and-conquer strategy, effectively breaking down intricate reasoning steps, accommodating diverse constraints, and managing multiple distinct sub-tasks in a well-organized manner. We further introduce an autonomous planning framework that completes the planning process by iteratively refining and expanding the hypertree-structured planning outlines. Experiments demonstrate the effectiveness of HTP, achieving state-of-the-art accuracy on the TravelPlanner benchmark with Gemini-1.5-Pro, resulting in a 3.6 times performance improvement over o1-preview.
Authors: Zirui Liu, Jiatong Li, Yan Zhuang, Qi Liu, Shuanghong Shen, Jie Ouyang, Mingyue Cheng, Shijin Wang
Abstract: Arena-based evaluation is a fundamental yet significant evaluation paradigm for modern AI models, especially large language models (LLMs). Existing framework based on ELO rating system suffers from the inevitable instability problem due to ranking inconsistency and the lack of attention to the varying abilities of annotators. In this paper, we introduce a novel stable arena framework to address these issues by enhancing the ELO Rating System. Specifically, we replace the iterative update method with a Maximum Likelihood Estimation (MLE) approach, m-ELO, and provide theoretical proof of the consistency and stability of the MLE approach for model ranking. Additionally, we proposed the am-ELO, which modify the Elo Rating's probability function to incorporate annotator abilities, enabling the simultaneous estimation of model scores and annotator reliability. Experiments demonstrate that this method ensures stability, proving that this framework offers a more robust, accurate, and stable evaluation method for LLMs.
Authors: Changzeng Fu, Zelin Fu, Qi Zhang, Xinhe Kuang, Jiacheng Dong, Kaifeng Su, Yikai Su, Wenbo Shi, Junfeng Yao, Yuliang Zhao, Shiqi Zhao, Jiadong Wang, Siyang Song, Chaoran Liu, Yuichiro Yoshikawa, Bj\"orn Schuller, Hiroshi Ishiguro
Abstract: Depression is a widespread mental health issue affecting diverse age groups, with notable prevalence among college students and the elderly. However, existing datasets and detection methods primarily focus on young adults, neglecting the broader age spectrum and individual differences that influence depression manifestation. Current approaches often establish a direct mapping between multimodal data and depression indicators, failing to capture the complexity and diversity of depression across individuals. This challenge includes two tracks based on age-specific subsets: Track 1 uses the MPDD-Elderly dataset for detecting depression in older adults, and Track 2 uses the MPDD-Young dataset for detecting depression in younger participants. The Multimodal Personality-aware Depression Detection (MPDD) Challenge aims to address this gap by incorporating multimodal data alongside individual difference factors. We provide a baseline model that fuses audio and video modalities with individual difference information to detect depression manifestations in diverse populations. This challenge aims to promote the development of more personalized and accurate de pression detection methods, advancing mental health research and fostering inclusive detection systems. More details are available on the official challenge website: https://hacilab.github.io/MPDDChallenge.github.io.
Authors: Junhao Zheng, Xidi Cai, Qiuke Li, Duzhen Zhang, ZhongZhi Li, Yingying Zhang, Le Song, Qianli Ma
Abstract: Lifelong learning is essential for intelligent agents operating in dynamic environments. Current large language model (LLM)-based agents, however, remain stateless and unable to accumulate or transfer knowledge over time. Existing benchmarks treat agents as static systems and fail to evaluate lifelong learning capabilities. We present LifelongAgentBench, the first unified benchmark designed to systematically assess the lifelong learning ability of LLM agents. It provides skill-grounded, interdependent tasks across three interactive environments, Database, Operating System, and Knowledge Graph, with automatic label verification, reproducibility, and modular extensibility. Extensive experiments reveal that conventional experience replay has limited effectiveness for LLM agents due to irrelevant information and context length constraints. We further introduce a group self-consistency mechanism that significantly improves lifelong learning performance. We hope LifelongAgentBench will advance the development of adaptive, memory-capable LLM agents.
Authors: Zidi Xiong, Shan Chen, Zhenting Qi, Himabindu Lakkaraju
Abstract: Large Reasoning Models (LRMs) have significantly enhanced their capabilities in complex problem-solving by introducing a thinking draft that enables multi-path Chain-of-Thought explorations before producing final answers. Ensuring the faithfulness of these intermediate reasoning processes is crucial for reliable monitoring, interpretation, and effective control. In this paper, we propose a systematic counterfactual intervention framework to rigorously evaluate thinking draft faithfulness. Our approach focuses on two complementary dimensions: (1) Intra-Draft Faithfulness, which assesses whether individual reasoning steps causally influence subsequent steps and the final draft conclusion through counterfactual step insertions; and (2) Draft-to-Answer Faithfulness, which evaluates whether final answers are logically consistent with and dependent on the thinking draft, by perturbing the draft's concluding logic. We conduct extensive experiments across six state-of-the-art LRMs. Our findings show that current LRMs demonstrate selective faithfulness to intermediate reasoning steps and frequently fail to faithfully align with the draft conclusions. These results underscore the need for more faithful and interpretable reasoning in advanced LRMs.
Authors: Hao Dong, Ziyue Qiao, Zhiyuan Ning, Qi Hao, Yi Du, Pengyang Wang, Yuanchun Zhou
Abstract: Temporal Knowledge Graphs (TKGs), as an extension of static Knowledge Graphs (KGs), incorporate the temporal feature to express the transience of knowledge by describing when facts occur. TKG extrapolation aims to infer possible future facts based on known history, which has garnered significant attention in recent years. Some existing methods treat TKG as a sequence of independent subgraphs to model temporal evolution patterns, demonstrating impressive reasoning performance. However, they still have limitations: 1) In modeling subgraph semantic evolution, they usually neglect the internal structural interactions between subgraphs, which are actually crucial for encoding TKGs. 2) They overlook the potential smooth features that do not lead to semantic changes, which should be distinguished from the semantic evolution process. Therefore, we propose a novel Disentangled Multi-span Evolutionary Network (DiMNet) for TKG reasoning. Specifically, we design a multi-span evolution strategy that captures local neighbor features while perceiving historical neighbor semantic information, thus enabling internal interactions between subgraphs during the evolution process. To maximize the capture of semantic change patterns, we design a disentangle component that adaptively separates nodes' active and stable features, used to dynamically control the influence of historical semantics on future evolution. Extensive experiments conducted on four real-world TKG datasets show that DiMNet demonstrates substantial performance in TKG reasoning, and outperforms the state-of-the-art up to 22.7% in MRR.
Authors: Hui Shen, Taiqiang Wu, Qi Han, Yunta Hsieh, Jizhou Wang, Yuyue Zhang, Yuxin Cheng, Zijian Hao, Yuansheng Ni, Xin Wang, Zhongwei Wan, Kai Zhang, Wendong Xu, Jing Xiong, Ping Luo, Wenhu Chen, Chaofan Tao, Zhuoqing Mao, Ngai Wong
Abstract: Existing benchmarks fail to capture a crucial aspect of intelligence: physical reasoning, the integrated ability to combine domain knowledge, symbolic reasoning, and understanding of real-world constraints. To address this gap, we introduce PhyX: the first large-scale benchmark designed to assess models capacity for physics-grounded reasoning in visual scenarios. PhyX includes 3K meticulously curated multimodal questions spanning 6 reasoning types across 25 sub-domains and 6 core physics domains: thermodynamics, electromagnetism, mechanics, modern physics, optics, and wave\&acoustics. In our comprehensive evaluation, even state-of-the-art models struggle significantly with physical reasoning. GPT-4o, Claude3.7-Sonnet, and GPT-o4-mini achieve only 32.5%, 42.2%, and 45.8% accuracy respectively-performance gaps exceeding 29% compared to human experts. Our analysis exposes critical limitations in current models: over-reliance on memorized disciplinary knowledge, excessive dependence on mathematical formulations, and surface-level visual pattern matching rather than genuine physical understanding. We provide in-depth analysis through fine-grained statistics, detailed case studies, and multiple evaluation paradigms to thoroughly examine physical reasoning capabilities. To ensure reproducibility, we implement a compatible evaluation protocol based on widely-used toolkits such as VLMEvalKit, enabling one-click evaluation. More details are available on our project page: https://phyx-bench.github.io/.
Authors: Sangyong Lee, Subo Hwang, Dohoon Kim
Abstract: In this paper, we propose MADCluster, a novel model-agnostic anomaly detection framework utilizing self-supervised clustering. MADCluster is applicable to various deep learning architectures and addresses the 'hypersphere collapse' problem inherent in existing deep learning-based anomaly detection methods. The core idea is to cluster normal pattern data into a 'single cluster' while simultaneously learning the cluster center and mapping data close to this center. Also, to improve expressiveness and enable effective single clustering, we propose a new 'One-directed Adaptive loss'. The optimization of this loss is mathematically proven. MADCluster consists of three main components: Base Embedder capturing high-dimensional temporal dynamics, Cluster Distance Mapping, and Sequence-wise Clustering for continuous center updates. Its model-agnostic characteristics are achieved by applying various architectures to the Base Embedder. Experiments on four time series benchmark datasets demonstrate that applying MADCluster improves the overall performance of comparative models. In conclusion, the compatibility of MADCluster shows potential for enhancing model performance across various architectures.
Authors: Licheng Pan, Yongqi Tong, Xin Zhang, Xiaolu Zhang, Jun Zhou, Zhixuan Chu
Abstract: Large language models (LLMs) have demonstrated remarkable capabilities across a wide range of tasks, yet they often refuse to answer legitimate queries-a phenomenon known as overrefusal. Overrefusal typically stems from over-conservative safety alignment, causing models to treat many reasonable prompts as potentially risky. To systematically understand this issue, we probe and leverage the models'safety decision boundaries to analyze and mitigate overrefusal. Our findings reveal that overrefusal is closely tied to misalignment at these boundary regions, where models struggle to distinguish subtle differences between benign and harmful content. Building on these insights, we present RASS, an automated framework for prompt generation and selection that strategically targets overrefusal prompts near the safety boundary. By harnessing steering vectors in the representation space, RASS efficiently identifies and curates boundary-aligned prompts, enabling more effective and targeted mitigation of overrefusal. This approach not only provides a more precise and interpretable view of model safety decisions but also seamlessly extends to multilingual scenarios.We have explored the safety decision boundaries of various LLMs and construct the MORBench evaluation set to facilitate robust assessment of model safety and helpfulness across multiple languages. Code and datasets will be released at https://anonymous.4open.science/r/RASS-80D3.
Authors: Ahmed Heakl, Yahia Salaheldin Shaaban, Martin Takac, Salem Lahlou, Zangir Iklassov
Abstract: Robust routing under uncertainty is central to real-world logistics, yet most benchmarks assume static, idealized settings. We present SVRPBench, the first open benchmark to capture high-fidelity stochastic dynamics in vehicle routing at urban scale. Spanning more than 500 instances with up to 1000 customers, it simulates realistic delivery conditions: time-dependent congestion, log-normal delays, probabilistic accidents, and empirically grounded time windows for residential and commercial clients. Our pipeline generates diverse, constraint-rich scenarios, including multi-depot and multi-vehicle setups. Benchmarking reveals that state-of-the-art RL solvers like POMO and AM degrade by over 20% under distributional shift, while classical and metaheuristic methods remain robust. To enable reproducible research, we release the dataset and evaluation suite. SVRPBench challenges the community to design solvers that generalize beyond synthetic assumptions and adapt to real-world uncertainty.
Authors: Keisuke Kawano, Takuro Kutsuna, Keisuke Sano
Abstract: Deep neural networks (DNNs) exhibit high performance in image recognition; however, the reasons for their strong generalization abilities remain unclear. A plausible hypothesis is that DNNs achieve robust and accurate predictions by identifying multiple pieces of evidence from images. Thus, to test this hypothesis, this study proposed minimal sufficient views (MSVs). MSVs is defined as a set of minimal regions within an input image that are sufficient to preserve the prediction of DNNs, thus representing the evidence discovered by the DNN. We empirically demonstrated a strong correlation between the number of MSVs (i.e., the number of pieces of evidence) and the generalization performance of the DNN models. Remarkably, this correlation was found to hold within a single DNN as well as between different DNNs, including convolutional and transformer models. This suggested that a DNN model that makes its prediction based on more evidence has a higher generalization performance. We proposed a metric based on MSVs for DNN model selection that did not require label information. Consequently, we empirically showed that the proposed metric was less dependent on the degree of overfitting, rendering it a more reliable indicator of model performance than existing metrics, such as average confidence.
Authors: Tianyu Chen, Haoyi Zhou, Ying Li, Hao Wang, Chonghan Gao, Rongye Shi, Shanghang Zhang, Jianxin Li
Abstract: Foundation models have revolutionized language modeling, while whether this success is replicated in scientific computing remains unexplored. We present OmniArch, the first prototype aiming at solving multi-scale and multi-physics scientific computing problems with physical alignment. We addressed all three challenges with one unified architecture. Its pre-training stage contains a Fourier Encoder-decoder fading out the disharmony across separated dimensions and a Transformer backbone integrating quantities through temporal dynamics, and the novel PDE-Aligner performs physics-informed fine-tuning under flexible conditions. As far as we know, we first conduct 1D-2D-3D united pre-training on the PDEBench, and it sets not only new performance benchmarks for 1D, 2D, and 3D PDEs but also demonstrates exceptional adaptability to new physics via in-context and zero-shot learning approaches, which supports realistic engineering applications and foresight physics discovery.
Authors: Filip Szatkowski, Yaoyue Zheng, Fei Yang, Bart{\l}omiej Twardowski, Tomasz Trzci\'nski, Joost van de Weijer
Abstract: Continual learning is crucial for applying machine learning in challenging, dynamic, and often resource-constrained environments. However, catastrophic forgetting - overwriting previously learned knowledge when new information is acquired - remains a major challenge. In this work, we examine the intermediate representations in neural network layers during continual learning and find that such representations are less prone to forgetting, highlighting their potential to accelerate computation. Motivated by these findings, we propose to use auxiliary classifiers(ACs) to enhance performance and demonstrate that integrating ACs into various continual learning methods consistently improves accuracy across diverse evaluation settings, yielding an average 10% relative gain. We also leverage the ACs to reduce the average cost of the inference by 10-60% without compromising accuracy, enabling the model to return the predictions before computing all the layers. Our approach provides a scalable and efficient solution for continual learning.
Authors: Shahin Atakishiyev, Mohammad Salameh, Randy Goebel
Abstract: The end-to-end learning pipeline is gradually creating a paradigm shift in the ongoing development of highly autonomous vehicles (AVs), largely due to advances in deep learning, the availability of large-scale training datasets, and improvements in integrated sensor devices. However, a lack of explainability in real-time decisions with contemporary learning methods impedes user trust and attenuates the widespread deployment and commercialization of such vehicles. Moreover, the issue is exacerbated when these vehicles are involved in or cause traffic accidents. Consequently, explainability in end-to-end autonomous driving is essential to build trust in vehicular automation. With that said, automotive researchers have not yet rigorously explored safety benefits and consequences of explanations in end-to-end autonomous driving. This paper aims to bridge the gaps between these topics and seeks to answer the following research question: What are safety implications of explanations in end-to-end autonomous driving? In this regard, we first revisit established safety and explainability concepts in end-to-end driving. Furthermore, we present critical case studies and show the pivotal role of explanations in enhancing driving safety. Finally, we describe insights from empirical studies and reveal potential value, limitations, and caveats of practical explainable AI methods with respect to their potential impacts on safety of end-to-end driving.
Authors: Yang Hu
Abstract: This research proposed a data-driven supply chain disruption response baseline framework based on intelligent recommender system technology as an initial SCRes reactive solution. To improve the data quality and reliability of the proposed IRS as a stable, secure, and resilient decision support system, blockchain technology is integrated into the baseline architecture. The smart contract is prototyped to demonstrate the information exchange mechanism under a BLC network environment. The BLC-IRS framework is then implemented with an industrial case to demonstrate its executable function. A system dynamics (SD) simulation model is adopted to validate the BLC-IRS framework as an effective digital SCRes enhancement measure. The simulation results indicated that the proposed BLC-IRS framework can be effectively implemented as a SC disruption mitigation measure in the SCRes response phase as reactive measure, enabling SC participants to react better to SC disruptions at the physical level. Compared to previous studies that limited at the conceptual level as the proactive SCRes measure with a standalone fashion, the developed BLC-IRS contributes an executable SCRes digital solution with synthetic technologies as a reactive SCRes measure for the SCRes community, by identifying the internal and external supplementary resource information in an agile, safe, and real-time manner after SC disruption.
Authors: Houxing Ren, Mingjie Zhan, Zhongyuan Wu, Aojun Zhou, Junting Pan, Hongsheng Li
Abstract: Code generation plays a crucial role in various tasks, such as code auto-completion and mathematical reasoning. Previous work has proposed numerous methods to enhance code generation performance, including integrating feedback from the compiler. Inspired by this, we present ReflectionCoder, a novel approach that effectively leverages reflection sequences constructed by integrating compiler feedback to improve one-off code generation performance. Furthermore, we propose reflection self-distillation and dynamically masked distillation to effectively utilize these reflection sequences. Extensive experiments on three benchmarks, i.e., HumanEval (+), MBPP (+), and MultiPL-E, demonstrate that models fine-tuned with our method achieve state-of-the-art performance. Beyond the code domain, we believe this approach can benefit other domains that focus on final results and require long reasoning paths. Code and data are available at https://github.com/SenseLLM/ReflectionCoder.
Authors: Oleksii Furman, Patryk Wielopolski, {\L}ukasz Lenkiewicz, Jerzy Stefanowski, Maciej Zi\k{e}ba
Abstract: The growing complexity of AI systems has intensified the need for transparency through Explainable AI (XAI). Counterfactual explanations (CFs) offer actionable "what-if" scenarios on three levels: Local CFs providing instance-specific insights, Global CFs addressing broader trends, and Group-wise CFs (GWCFs) striking a balance and revealing patterns within cohesive groups. Despite the availability of methods for each granularity level, the field lacks a unified method that integrates these complementary approaches. We address this limitation by proposing a gradient-based optimization method for differentiable models that generates Local, Global, and Group-wise Counterfactual Explanations in a unified manner. We especially enhance GWCF generation by combining instance grouping and counterfactual generation into a single efficient process, replacing traditional two-step methods. Moreover, to ensure trustworthiness, we innovatively introduce the integration of plausibility criteria into the GWCF domain, making explanations both valid and realistic. Our results demonstrate the method's effectiveness in balancing validity, proximity, and plausibility while optimizing group granularity, with practical utility validated through practical use cases.
Authors: Jiangkai Wu, Liming Liu, Yunpeng Tan, Junlin Hao, Xinggong Zhang
Abstract: With the exponential growth of video traffic, traditional video streaming systems are approaching their limits in compression efficiency and communication capacity. To further reduce bitrate while maintaining quality, we propose Promptus, a disruptive semantic communication system that streaming prompts instead of video content, which represents real-world video frames with a series of "prompts" for delivery and employs Stable Diffusion to generate videos at the receiver. To ensure that the generated video is pixel-aligned with the original video, a gradient descent-based prompt fitting framework is proposed. Further, a low-rank decomposition-based bitrate control algorithm is introduced to achieve adaptive bitrate. For inter-frame compression, an interpolation-aware fitting algorithm is proposed. Evaluations across various video genres demonstrate that, compared to H.265, Promptus can achieve more than a 4x bandwidth reduction while preserving the same perceptual quality. On the other hand, at extremely low bitrates, Promptus can enhance the perceptual quality by 0.139 and 0.118 (in LPIPS) compared to VAE and H.265, respectively, and decreases the ratio of severely distorted frames by 89.3% and 91.7%. Our work opens up a new paradigm for efficient video communication. Promptus is open-sourced at: https://github.com/JiangkaiWu/Promptus.
Authors: Yuansan Liu, Sudanthi Wijewickrema, Dongting Hu, Christofer Bester, Stephen O'Leary, James Bailey
Abstract: Recent innovations in diffusion probabilistic models have paved the way for significant progress in image, text and audio generation, leading to their applications in generative time series forecasting. However, leveraging such abilities to model highly stochastic time series data remains a challenge. In this paper, we propose a novel Stochastic Diffusion (StochDiff) model which learns data-driven prior knowledge at each time step by utilizing the representational power of the stochastic latent spaces to model the variability of the multivariate time series data. The learnt prior knowledge helps the model to capture complex temporal dynamics and the inherent uncertainty of the data. This improves its ability to model highly stochastic time series data. Through extensive experiments on real-world datasets, we demonstrate the effectiveness of our proposed model on stochastic time series forecasting. Additionally, we showcase an application of our model for real-world surgical guidance, highlighting its potential to benefit the medical community.
Authors: Niamh Belton, Aonghus Lawlor, Kathleen M. Curran
Abstract: The diagnostic accuracy and subjectivity of existing Knee Osteoarthritis (OA) ordinal grading systems has been a subject of on-going debate and concern. Existing automated solutions are trained to emulate these imperfect systems, whilst also being reliant on large annotated databases for fully-supervised training. This work proposes a three stage approach for automated continuous grading of knee OA that is built upon the principles of Anomaly Detection (AD); learning a robust representation of healthy knee X-rays and grading disease severity based on its distance to the centre of normality. In the first stage, SS-FewSOME is proposed, a self-supervised AD technique that learns the 'normal' representation, requiring only examples of healthy subjects and <3% of the labels that existing methods require. In the second stage, this model is used to pseudo label a subset of unlabelled data as 'normal' or 'anomalous', followed by denoising of pseudo labels with CLIP. The final stage involves retraining on labelled and pseudo labelled data using the proposed Dual Centre Representation Learning (DCRL) which learns the centres of two representation spaces; normal and anomalous. Disease severity is then graded based on the distance to the learned centres. The proposed methodology outperforms existing techniques by margins of up to 24% in terms of OA detection and the disease severity scores correlate with the Kellgren-Lawrence grading system at the same level as human expert performance. Code available at https://github.com/niamhbelton/SS-FewSOME_Disease_Severity_Knee_Osteoarthritis.
URLs: https://github.com/niamhbelton/SS-FewSOME_Disease_Severity_Knee_Osteoarthritis.
Authors: Yuxiang Chai, Siyuan Huang, Yazhe Niu, Han Xiao, Liang Liu, Dingyu Zhang, Shuai Ren, Hongsheng Li
Abstract: AI agents have drawn increasing attention mostly on their ability to perceive environments, understand tasks, and autonomously achieve goals. To advance research on AI agents in mobile scenarios, we introduce the Android Multi-annotation EXpo (AMEX), a comprehensive, large-scale dataset designed for generalist mobile GUI-control agents which are capable of completing tasks by directly interacting with the graphical user interface (GUI) on mobile devices. AMEX comprises over 104K high-resolution screenshots from popular mobile applications, which are annotated at multiple levels. Unlike existing GUI-related datasets, e.g., Rico, AitW, etc., AMEX includes three levels of annotations: GUI interactive element grounding, GUI screen and element functionality descriptions, and complex natural language instructions with stepwise GUI-action chains. We develop this dataset from a more instructive and detailed perspective, complementing the general settings of existing datasets. Additionally, we finetune a baseline model SPHINX Agent and illustrate the effectiveness of AMEX.The project is available at https://yxchai.com/AMEX/.
Authors: Weiqi Wu, Hongqiu Wu, Hai Zhao
Abstract: The Turing test examines whether AIs exhibit human-like behaviour in natural language conversations. The traditional setting limits each participant to one message at a time and requires constant human participation. This fails to reflect a natural conversational style and hinders the evaluation of dialogue agents based on Large Language Models (LLMs) in complex and prolonged interactions. This paper proposes \textbf{\textsc{X-Turing}}, which enhances the original test with a \textit{burst dialogue} pattern, allowing more dynamic exchanges using consecutive messages. It further reduces human workload by iteratively generating dialogues that simulate the long-term interaction between the agent and a human to compose the majority of the test process. With the \textit{pseudo-dialogue} history, the agent then engages in a shorter dialogue with a real human, which is paired with a human-human conversation on the same topic to be judged using questionnaires. We introduce the \textit{X-Turn Pass-Rate} metric to assess the human likeness of LLMs across varying durations. While LLMs like GPT-4 initially perform well, achieving pass rates of 51.9\% and 38.9\% during 3 turns and 10 turns of dialogues respectively, their performance drops as the dialogue progresses, which underscores the difficulty in maintaining consistency in the long term.
Authors: Arun Verma, Indrajit Saha, Makoto Yokoo, Bryan Kian Hsiang Low
Abstract: This paper considers a novel variant of the online fair division problem involving multiple agents in which a learner sequentially observes an indivisible item that has to be irrevocably allocated to one of the agents while satisfying a fairness and efficiency constraint. Existing algorithms assume a small number of items with a sufficiently large number of copies, which ensures a good utility estimation for all item-agent pairs from noisy bandit feedback. However, this assumption may not hold in many real-life applications, for example, an online platform that has a large number of users (items) who use the platform's service providers (agents) only a few times (a few copies of items), which makes it difficult to accurately estimate utilities for all item-agent pairs. To address this, we assume utility is an unknown function of item-agent features. We then propose algorithms that model online fair division as a contextual bandit problem, with sub-linear regret guarantees. Our experimental results further validate the effectiveness of the proposed algorithms.
Authors: Chuheng Zhang, Wei Shen, Li Zhao, Xuyun Zhang, Xiaolong Xu, Wanchun Dou, Jiang Biang
Abstract: While direct policy optimization methods exist, pioneering LLMs are fine-tuned with reinforcement learning from human feedback (RLHF) to generate better responses under the supervision of a reward model learned from preference data. One major challenge of RLHF is the inaccuracy of the intermediate reward model, especially in the tasks that requires complex reasoning for the reward model to score a response. We find that the reliability of the reward model varies across responses assigned with different rewards. This motivates us to filter the samples whose rewards may be unreliable to improve the signal-to-noise ratio during policy learning, resulting in Policy Filtration for Proximal Policy Optimization (PF-PPO). To choose a proper policy filtering strategy, we use the coefficient of determination (R2) between the rewards and actual scores on filtered samples as the metrics to help us find promising strategies since it measures how well the rewards filtered by PF-PPO indicate real performance. We provide extensive experiments to validate the effectiveness of PF-PPO in code generation and math reasoning tasks. In code generation, PF-PPO achieves the state-of-the-art performance of 7-billion-parameter models on HumanEval (+7.9%), MBPP (+0.7%), and LeetCode Contest (+10.0%) which is a more challenging benchmark created by us. In math reasoning, PF-PPO yields performance increase using different reward models and benchmarks (Ape210K and CMATH). Code is available on https://github.com/DtYXs/verl/tree/pf-ppo.
Authors: Wenhui Diao, Haichen Yu, Kaiyue Kang, Tong Ling, Di Liu, Yingchao Feng, Hanbo Bi, Libo Ren, Xuexue Li, Yongqiang Mao, Xian Sun
Abstract: Aerial Remote Sensing (ARS) vision tasks pose significant challenges due to the unique characteristics of their viewing angles. Existing research has primarily focused on algorithms for specific tasks, which have limited applicability in a broad range of ARS vision applications. This paper proposes the RingMo-Aerial model, aiming to fill the gap in foundation model research in the field of ARS vision. By introducing the Frequency-Enhanced Multi-Head Self-Attention (FE-MSA) mechanism and an affine transformation-based contrastive learning pre-training method, the model's detection capability for small targets is enhanced and optimized for the tilted viewing angles characteristic of ARS. Furthermore, the ARS-Adapter, an efficient parameter fine-tuning method, is proposed to improve the model's adaptability and effectiveness in various ARS vision tasks. Experimental results demonstrate that RingMo-Aerial achieves SOTA performance on multiple downstream tasks. This indicates the practicality and efficacy of RingMo-Aerial in enhancing the performance of ARS vision tasks.
Authors: Ruifeng Ren, Zhicong Li, Yong Liu
Abstract: Transformers have become the backbone of modern Large Language Models (LLMs); however, their inference overhead grows linearly with the sequence length, posing challenges for modeling long sequences. In light of this, Mamba has attracted attention for maintaining a constant inference size, with empirical evidence demonstrating that it can match Transformer performance in sequence modeling while significantly reducing computational costs. However, an open question remains: can Mamba always bring savings while achieving performance comparable to Transformers? In this paper, we focus on analyzing the expressive ability of Mamba to perform our defined COPY operation and Chain of Thought (CoT) reasoning. First, inspired by the connection between Mamba and linear attention, we show that constant-sized Mamba may struggle to perform COPY operations while Transformers can handle them more easily. However, when the size of Mamba grows linearly with the input sequence length, it can accurately perform COPY, but in this case, Mamba no longer provides overhead savings. Based on this observation, we further analyze Mamba's ability to tackle CoT tasks, which can be described by the Dynamic Programming (DP) problems. Our findings suggest that to solve arbitrary DP problems, the total cost of Mamba is still comparable to standard Transformers. However, similar to efficient Transformers, when facing DP problems with favorable properties such as locality, Mamba can provide savings in overhead. Our experiments on the copy and CoT tasks further demonstrate Mamba's limitations compared to Transformers in learning these tasks.
Authors: Long Zhao, Sanghyun Woo, Ziyu Wan, Yandong Li, Han Zhang, Boqing Gong, Hartwig Adam, Xuhui Jia, Ting Liu
Abstract: In generative modeling, tokenization simplifies complex data into compact, structured representations, creating a more efficient, learnable space. For high-dimensional visual data, it reduces redundancy and emphasizes key features for high-quality generation. Current visual tokenization methods rely on a traditional autoencoder framework, where the encoder compresses data into latent representations, and the decoder reconstructs the original input. In this work, we offer a new perspective by proposing denoising as decoding, shifting from single-step reconstruction to iterative refinement. Specifically, we replace the decoder with a diffusion process that iteratively refines noise to recover the original image, guided by the latents provided by the encoder. We evaluate our approach by assessing both reconstruction (rFID) and generation quality (FID), comparing it to state-of-the-art autoencoding approaches. By adopting iterative reconstruction through diffusion, our autoencoder, namely Epsilon-VAE, achieves high reconstruction quality, which in turn enhances downstream generation quality by 22% at the same compression rates or provides 2.3x inference speedup through increasing compression rates. We hope this work offers new insights into integrating iterative generation and autoencoding for improved compression and generation.
Authors: Zeyu Zhang, Sixu Yan, Muzhi Han, Zaijin Wang, Xinggang Wang, Song-Chun Zhu, Hangxin Liu
Abstract: We propose M3Bench, a new benchmark for whole-body motion generation in mobile manipulation tasks. Given a 3D scene context, M3Bench requires an embodied agent to reason about its configuration, environmental constraints, and task objectives to generate coordinated whole-body motion trajectories for object rearrangement. M3Bench features 30,000 object rearrangement tasks across 119 diverse scenes, providing expert demonstrations generated by our newly developed M3BenchMaker, an automatic data generation tool that produces whole-body motion trajectories from high-level task instructions using only basic scene and robot information. Our benchmark includes various task splits to evaluate generalization across different dimensions and leverages realistic physics simulation for trajectory assessment. Extensive evaluation analysis reveals that state-of-the-art models struggle with coordinating base-arm motion while adhering to environmental and task-specific constraints, underscoring the need for new models to bridge this gap. By releasing M3Bench and M3BenchMaker we aim to advance robotics research toward more adaptive and capable mobile manipulation in diverse, real-world environments.
Authors: Ariel Flint Ashery, Luca Maria Aiello, Andrea Baronchelli
Abstract: Social conventions are the backbone of social coordination, shaping how individuals form a group. As growing populations of artificial intelligence (AI) agents communicate through natural language, a fundamental question is whether they can bootstrap the foundations of a society. Here, we present experimental results that demonstrate the spontaneous emergence of universally adopted social conventions in decentralized populations of large language model (LLM) agents. We then show how strong collective biases can emerge during this process, even when agents exhibit no bias individually. Last, we examine how committed minority groups of adversarial LLM agents can drive social change by imposing alternative social conventions on the larger population. Our results show that AI systems can autonomously develop social conventions without explicit programming and have implications for designing AI systems that align, and remain aligned, with human values and societal goals.
Authors: Mohammad Mozaffari, Amir Yazdanbakhsh, Maryam Mehri Dehnavi
Abstract: Conventional model compression techniques for LLMs address high memory consumption and slow inference challenges but typically require computationally expensive retraining to preserve accuracy. In contrast, one-shot compression methods eliminate retraining cost, but struggle to achieve accuracy comparable to dense models. This paper presents SLIM, a new one-shot compression framework that holistically integrates hardware-friendly quantization, sparsity, and low-rank approximation into a unified process. First, we formulate the quantization process using a probabilistic approach (SLIM-Quant) that enables us to apply uniform quantization. Then, we use an existing one-shot pruning method to apply semi-structured sparsity on top of the quantized weights. Finally, to compensate for the introduced aggregated quantization and sparsity error, we use a novel saliency function with unique invertible and additive features that enables us to mathematically compute the value of low-rank adapters. SLIM improves model accuracy by up to 5.66% (LLaMA-2-7B) for 2:4 sparsity with 4-bit weight quantization, outperforming prior methods. Models compressed with SLIM achieve up to 4.3x and 3.8x on Nvidia RTX3060 and A100 GPUs, respectively. Additionally, they achieve up to 0.23x end-to-end memory reduction in comparison to their dense counterparts. We also propose an optional PEFT recipe that further improves accuracy by up to 1.66% (LLaMA-2-13B) compared to SLIM without fine-tuning.
Authors: Hojoon Lee, Dongyoon Hwang, Donghu Kim, Hyunseung Kim, Jun Jet Tai, Kaushik Subramanian, Peter R. Wurman, Jaegul Choo, Peter Stone, Takuma Seno
Abstract: Recent advances in CV and NLP have been largely driven by scaling up the number of network parameters, despite traditional theories suggesting that larger networks are prone to overfitting. These large networks avoid overfitting by integrating components that induce a simplicity bias, guiding models toward simple and generalizable solutions. However, in deep RL, designing and scaling up networks have been less explored. Motivated by this opportunity, we present SimBa, an architecture designed to scale up parameters in deep RL by injecting a simplicity bias. SimBa consists of three components: (i) an observation normalization layer that standardizes inputs with running statistics, (ii) a residual feedforward block to provide a linear pathway from the input to output, and (iii) a layer normalization to control feature magnitudes. By scaling up parameters with SimBa, the sample efficiency of various deep RL algorithms-including off-policy, on-policy, and unsupervised methods-is consistently improved. Moreover, solely by integrating SimBa architecture into SAC, it matches or surpasses state-of-the-art deep RL methods with high computational efficiency across DMC, MyoSuite, and HumanoidBench. These results demonstrate SimBa's broad applicability and effectiveness across diverse RL algorithms and environments.
Authors: Haozhen Zhang, Tao Feng, Jiaxuan You
Abstract: Retrieval-augmented generation (RAG) has revitalized Large Language Models (LLMs) by injecting non-parametric factual knowledge. Compared with long-context LLMs, RAG is considered an effective summarization tool in a more concise and lightweight manner, which can interact with LLMs multiple times using diverse queries to get comprehensive responses. However, the LLM-generated historical responses, which contain potentially insightful information, are largely neglected and discarded by existing approaches, leading to suboptimal results. In this paper, we propose $\textit{graph of records}$ ($\textbf{GoR}$), which leverages historical responses generated by LLMs to enhance RAG for long-context global summarization. Inspired by the $\textit{retrieve-then-generate}$ paradigm of RAG, we construct a graph by establishing an edge between the retrieved text chunks and the corresponding LLM-generated response. To further uncover the intricate correlations between them, GoR features a $\textit{graph neural network}$ and an elaborately designed $\textit{BERTScore}$-based objective for self-supervised model training, enabling seamless supervision signal backpropagation between reference summaries and node embeddings. We comprehensively compare GoR with 12 baselines across four long-context summarization datasets, and the results indicate that our proposed method reaches the best performance ($\textit{e.g.}$, 15%, 8%, and 19% improvement over retrievers w.r.t. Rouge-L, Rouge-1, and Rouge-2 on the WCEP dataset). Extensive experiments further demonstrate the effectiveness of GoR.
Authors: Anjiang Wei, Allen Nie, Thiago S. F. X. Teixeira, Rohan Yadav, Wonchan Lee, Ke Wang, Alex Aiken
Abstract: Modern scientific discovery increasingly relies on high-performance computing for complex modeling and simulation. A key challenge in improving parallel program performance is efficiently mapping tasks to processors and data to memory, a process dictated by intricate, low-level system code known as mappers. Developing high-performance mappers demands days of manual tuning, posing a significant barrier for domain scientists without systems expertise. We introduce a framework that automates mapper development with generative optimization, leveraging richer feedback beyond scalar performance metrics. Our approach features the Agent-System Interface, which includes a Domain-Specific Language (DSL) to abstract away the low-level complexity of system code and define a structured search space, as well as AutoGuide, a mechanism that interprets raw execution output into actionable feedback. Unlike traditional reinforcement learning methods such as OpenTuner, which rely solely on scalar feedback, our method finds superior mappers in far fewer iterations. With just 10 iterations, it outperforms OpenTuner even after 1000 iterations, achieving 3.8X faster performance. Our approach finds mappers that surpass expert-written mappers by up to 1.34X speedup across nine benchmarks while reducing tuning time from days to minutes.
Authors: Yaoqi Guo, Zhenpeng Chen, Jie M. Zhang, Yang Liu, Yun Ma
Abstract: Code generation, the automatic creation of source code from natural language descriptions, has garnered significant attention due to its potential to streamline software development. Inspired by research that links task-personality alignment with improved development outcomes, we conduct an empirical study on personality-guided code generation using large language models (LLMs). Specifically, we investigate how emulating personality traits appropriate to the coding tasks affects LLM performance. We extensively evaluate this approach using seven widely adopted LLMs across four representative datasets. Our results show that personality guidance significantly enhances code generation accuracy, with improved pass rates in 23 out of 28 LLM-dataset combinations. Notably, in 11 cases, the improvement exceeds 5%, and in 5 instances, it surpasses 10%, with the highest gain reaching 12.9%. Additionally, personality guidance can be easily integrated with other prompting strategies to further boost performance. We open-source our code and data at https://github.com/IanWalls/Persona-Code.
Authors: Yihua Shao, Yan Gu, Siyu Chen, Haiyang Liu, Zixian Zhu, Zijian Ling, Minxi Yan, Ziyang Yan, Chenyu Zhang, Michele Magno, Haotong Qin, Yan Wang, Jingcai Guo, Ling Shao, Hao Tang
Abstract: Large language models (LLMs) show impressive performance in solving complex language tasks. However, its large number of parameters presents significant challenges for the deployment. So, compressing LLMs to low bits can enable to deploy on resource-constrained devices. To address this problem, we propose gradient-aware weight quantization (GWQ), the first quantization approach for low-bit weight quantization that leverages gradients to localize outliers, requiring only a minimal amount of calibration data for outlier detection. GWQ retains the top 1\% outliers preferentially at FP16 precision, while the remaining non-outlier weights are stored in a low-bit. We widely evaluate GWQ on different task include language modeling, grounding detection, massive multitask language understanding and vision-language question and answering. Results show that models quantified by GWQ performs better than other quantization method. During quantization process, GWQ only need one calibration set to realize effective quant. Also, GWQ achieves 1.2x inference speedup in comparison to the original model and effectively reduces the inference memory.
Authors: Bijoy Ahmed Saiem, MD Sadik Hossain Shanto, Rakib Ahsan, Md Rafi ur Rashid
Abstract: As the integration of the Large Language Models (LLMs) into various applications increases, so does their susceptibility to misuse, raising significant security concerns. Numerous jailbreak attacks have been proposed to assess the security defense of LLMs. Current jailbreak attacks mainly rely on scenario camouflage, prompt obfuscation, prompt optimization, and prompt iterative optimization to conceal malicious prompts. In particular, sequential prompt chains in a single query can lead LLMs to focus on certain prompts while ignoring others, facilitating context manipulation. This paper introduces SequentialBreak, a novel jailbreak attack that exploits this vulnerability. We discuss several scenarios, not limited to examples like Question Bank, Dialog Completion, and Game Environment, where the harmful prompt is embedded within benign ones that can fool LLMs into generating harmful responses. The distinct narrative structures of these scenarios show that SequentialBreak is flexible enough to adapt to various prompt formats beyond those discussed. Extensive experiments demonstrate that SequentialBreak uses only a single query to achieve a substantial gain of attack success rate over existing baselines against both open-source and closed-source models. Through our research, we highlight the urgent need for more robust and resilient safeguards to enhance LLM security and prevent potential misuse. All the result files and website associated with this research are available in this GitHub repository: https://anonymous.4open.science/r/JailBreakAttack-4F3B/.
URLs: https://anonymous.4open.science/r/JailBreakAttack-4F3B/.
Authors: Guoguo Ai, Guansong Pang, Hezhe Qiao, Yuan Gao, Hui Yan
Abstract: Graph Transformers (GTs) have demonstrated remarkable performance in graph representation learning over popular graph neural networks (GNNs). However, self--attention, the core module of GTs, preserves only low-frequency signals in graph features, leading to ineffectiveness in capturing other important signals like high-frequency ones. Some recent GT models help alleviate this issue, but their flexibility and expressiveness are still limited since the filters they learn are fixed on predefined graph spectrum or spectral order. To tackle this challenge, we propose a Graph Fourier Kolmogorov-Arnold Transformer (GrokFormer), a novel GT model that learns highly expressive spectral filters with adaptive graph spectrum and spectral order through a Fourier series modeling over learnable activation functions. We demonstrate theoretically and empirically that the proposed GrokFormer filter offers better expressiveness than other spectral methods. Comprehensive experiments on 10 real-world node classification datasets across various domains, scales, and graph properties, as well as 5 graph classification datasets, show that GrokFormer outperforms state-of-the-art GTs and GNNs. Our code is available at https://github.com/GGA23/GrokFormer
Authors: Aladin Djuhera, Amin Seffo, Vlad C. Andrei, Holger Boche, Walid Saad
Abstract: Path planning under wireless performance constraints is a complex challenge in robot navigation. However, naively incorporating such constraints into classical planning algorithms often incurs prohibitive search costs. In this paper, we propose SCoTT, a wireless-aware path planning framework that leverages vision-language models (VLMs) to co-optimize average path gains and trajectory length using wireless heatmap images and ray-tracing data from a digital twin (DT). At the core of our framework is Strategic Chain-of-Thought Tasking (SCoTT), a novel prompting paradigm that decomposes the exhaustive search problem into structured subtasks, each solved via chain-of-thought prompting. To establish strong baselines, we compare classical A* and wireless-aware extensions of it, and derive DP-WA*, an optimal, iterative dynamic programming algorithm that incorporates all path gains and distance metrics from the DT, but at significant computational cost. In extensive experiments, we show that SCoTT achieves path gains within 2% of DP-WA* while consistently generating shorter trajectories. Moreover, SCoTT's intermediate outputs can be used to accelerate DP-WA* by reducing its search space, saving up to 62% in execution time. We validate our framework using four VLMs, demonstrating effectiveness across both large and small models, thus making it applicable to a wide range of compact models at low inference cost. We also show the practical viability of our approach by deploying SCoTT as a ROS node within Gazebo simulations. Finally, we discuss data acquisition pipelines, compute requirements, and deployment considerations for VLMs in 6G-enabled DTs, underscoring the potential of natural language interfaces for wireless-aware navigation in real-world applications.
Authors: An T. Le, Kay Hansel, Jo\~ao Carvalho, Joe Watson, Julen Urain, Armin Biess, Georgia Chalvatzaki, Jan Peters
Abstract: Batch planning is increasingly necessary to quickly produce diverse and quality motion plans for downstream learning applications, such as distillation and imitation learning. This paper presents Global Tensor Motion Planning (GTMP) -- a sampling-based motion planning algorithm comprising only tensor operations. We introduce a novel discretization structure represented as a random multipartite graph, enabling efficient vectorized sampling, collision checking, and search. We provide a theoretical investigation showing that GTMP exhibits probabilistic completeness while supporting modern GPU/TPU. Additionally, by incorporating smooth structures into the multipartite graph, GTMP directly plans smooth splines without requiring gradient-based optimization. Experiments on lidar-scanned occupancy maps and the MotionBenchMarker dataset demonstrate GTMP's computation efficiency in batch planning compared to baselines, underscoring GTMP's potential as a robust, scalable planner for diverse applications and large-scale robot learning tasks.
Authors: Jingzhe Liu, Haitao Mao, Zhikai Chen, Bingheng Li, Wenqi Fan, Mingxuan Ju, Tong Zhao, Neil Shah, Jiliang Tang
Abstract: Graph Neural Networks (GNNs) have emerged as a powerful tool to capture intricate network patterns, achieving success across different domains. However, existing GNNs require careful domain-specific architecture designs and training from scratch on each dataset, leading to an expertise-intensive process with difficulty in generalizing across graphs from different domains. Therefore, it can be hard for practitioners to infer which GNN model can generalize well to graphs from their domains. To address this challenge, we propose a novel cross-domain pretraining framework, "one model for one graph," which overcomes the limitations of previous approaches that failed to use a single GNN to capture diverse graph patterns across domains with significant gaps. Specifically, we pretrain a bank of expert models, with each one corresponding to a specific dataset. When inferring to a new graph, gating functions choose a subset of experts to effectively integrate prior model knowledge while avoiding negative transfer. Extensive experiments consistently demonstrate the superiority of our proposed method on both link prediction and node classification tasks.
Authors: Yuyang Wang, Anurag Ranjan, Josh Susskind, Miguel Angel Bautista
Abstract: Flow matching models have emerged as a powerful method for generative modeling on domains like images or videos, and even on irregular or unstructured data like 3D point clouds or even protein structures. These models are commonly trained in two stages: first, a data compressor is trained, and in a subsequent training stage a flow matching generative model is trained in the latent space of the data compressor. This two-stage paradigm sets obstacles for unifying models across data domains, as hand-crafted compressors architectures are used for different data modalities. To this end, we introduce INRFlow, a domain-agnostic approach to learn flow matching transformers directly in ambient space. Drawing inspiration from INRs, we introduce a conditionally independent point-wise training objective that enables INRFlow to make predictions continuously in coordinate space. Our empirical results demonstrate that INRFlow effectively handles different data modalities such as images, 3D point clouds and protein structure data, achieving strong performance in different domains and outperforming comparable approaches. INRFlow is a promising step towards domain-agnostic flow matching generative models that can be trivially adopted in different data domains.
Authors: Xiangnan Yu, Hao Xu, Zhiping Mao, HongGuang Sun, Yong Zhang, Dongxiao Zhang, Yuntian Chen
Abstract: In complex physical systems, conventional differential equations often fall short in capturing non-local and memory effects, as they are limited to local dynamics and integer-order interactions. This study introduces a stepwise data-driven framework for discovering fractional differential equations (FDEs) directly from data. FDEs, known for their capacity to model non-local dynamics with fewer parameters than integer-order derivatives, can represent complex systems with long-range interactions. Our framework applies deep neural networks as surrogate models for denoising and reconstructing sparse and noisy observations while using Gaussian-Jacobi quadrature to handle the challenges posed by singularities in fractional derivatives. To optimize both the sparse coefficients and fractional order, we employ an alternating optimization approach that combines sparse regression with global optimization techniques. We validate the framework across various datasets, including synthetic anomalous diffusion data, experimental data on the creep behavior of frozen soils, and single-particle trajectories modeled by L\'{e}vy motion. Results demonstrate the framework's robustness in identifying the structure of FDEs across diverse noise levels and its capacity to capture integer-order dynamics, offering a flexible approach for modeling memory effects in complex systems.
Authors: Mohit Chandra, Suchismita Naik, Denae Ford, Ebele Okoli, Munmun De Choudhury, Mahsa Ershadi, Gonzalo Ramos, Javier Hernandez, Ananya Bhattacharjee, Shahed Warreth, Jina Suh
Abstract: Recent gains in popularity of AI conversational agents have led to their increased use for improving productivity and supporting well-being. While previous research has aimed to understand the risks associated with interactions with AI conversational agents, these studies often fall short in capturing the lived experiences of individuals. Additionally, psychological risks have often been presented as a sub-category within broader AI-related risks in past taxonomy works, leading to under-representation of the impact of psychological risks of AI use. To address these challenges, our work presents a novel risk taxonomy focusing on psychological risks of using AI gathered through the lived experiences of individuals. We employed a mixed-method approach, involving a comprehensive survey with 283 people with lived mental health experience and workshops involving experts with lived experience to develop a psychological risk taxonomy. Our taxonomy features 19 AI behaviors, 21 negative psychological impacts, and 15 contexts related to individuals. Additionally, we propose a novel multi-path vignette-based framework for understanding the complex interplay between AI behaviors, psychological impacts, and individual user contexts. Finally, based on the feedback obtained from the workshop sessions, we present design recommendations for developing safer and more robust AI agents. Our work offers an in-depth understanding of the psychological risks associated with AI conversational agents and provides actionable recommendations for policymakers, researchers, and developers.
Authors: Taeho Hwang, Sukmin Cho, Soyeong Jeong, Hoyun Song, SeungYoon Han, Jong C. Park
Abstract: We introduce EXIT, an extractive context compression framework that enhances both the effectiveness and efficiency of retrieval-augmented generation (RAG) in question answering (QA). Current RAG systems often struggle when retrieval models fail to rank the most relevant documents, leading to the inclusion of more context at the expense of latency and accuracy. While abstractive compression methods can drastically reduce token counts, their token-by-token generation process significantly increases end-to-end latency. Conversely, existing extractive methods reduce latency but rely on independent, non-adaptive sentence selection, failing to fully utilize contextual information. EXIT addresses these limitations by classifying sentences from retrieved documents - while preserving their contextual dependencies - enabling parallelizable, context-aware extraction that adapts to query complexity and retrieval quality. Our evaluations on both single-hop and multi-hop QA tasks show that EXIT consistently surpasses existing compression methods and even uncompressed baselines in QA accuracy, while also delivering substantial reductions in inference time and token count. By improving both effectiveness and efficiency, EXIT provides a promising direction for developing scalable, high-quality QA solutions in RAG pipelines. Our code is available at https://github.com/ThisIsHwang/EXIT
Authors: Yuzheng Cai, Zhenyue Guo, Yiwen Pei, Wanrui Bian, Weiguo Zheng
Abstract: Recent advancements in large language models (LLMs) have shown impressive versatility across various tasks. To eliminate their hallucinations, retrieval-augmented generation (RAG) has emerged as a powerful approach, leveraging external knowledge sources like knowledge graphs (KGs). In this paper, we study the task of KG-driven RAG and propose a novel Similar Graph Enhanced Retrieval-Augmented Generation (SimGRAG) method. It effectively addresses the challenge of aligning query texts and KG structures through a two-stage process: (1) query-to-pattern, which uses an LLM to transform queries into a desired graph pattern, and (2) pattern-to-subgraph, which quantifies the alignment between the pattern and candidate subgraphs using a graph semantic distance (GSD) metric. We also develop an optimized retrieval algorithm that efficiently identifies the top-k subgraphs within 1-second on a 10-million-scale KG. Extensive experiments show that SimGRAG outperforms state-of-the-art KG-driven RAG methods in both question answering and fact verification. Our code is available at https://github.com/YZ-Cai/SimGRAG.
Authors: Nilanjana Das, Edward Raff, Aman Chadha, Manas Gaur
Abstract: As the AI systems become deeply embedded in social media platforms, we've uncovered a concerning security vulnerability that goes beyond traditional adversarial attacks. It becomes important to assess the risks of LLMs before the general public use them on social media platforms to avoid any adverse impacts. Unlike obvious nonsensical text strings that safety systems can easily catch, our work reveals that human-readable situation-driven adversarial full-prompts that leverage situational context are effective but much harder to detect. We found that skilled attackers can exploit the vulnerabilities in open-source and proprietary LLMs to make a malicious user query safe for LLMs, resulting in generating a harmful response. This raises an important question about the vulnerabilities of LLMs. To measure the robustness against human-readable attacks, which now present a potent threat, our research makes three major contributions. First, we developed attacks that use movie scripts as situational contextual frameworks, creating natural-looking full-prompts that trick LLMs into generating harmful content. Second, we developed a method to transform gibberish adversarial text into readable, innocuous content that still exploits vulnerabilities when used within the full-prompts. Finally, we enhanced the AdvPrompter framework with p-nucleus sampling to generate diverse human-readable adversarial texts that significantly improve attack effectiveness against models like GPT-3.5-Turbo-0125 and Gemma-7b. Our findings show that these systems can be manipulated to operate beyond their intended ethical boundaries when presented with seemingly normal prompts that contain hidden adversarial elements. By identifying these vulnerabilities, we aim to drive the development of more robust safety mechanisms that can withstand sophisticated attacks in real-world applications.
Authors: Lovisa Hagstr\"om, Sara Vera Marjanovi\'c, Haeun Yu, Arnav Arora, Christina Lioma, Maria Maistro, Pepa Atanasova, Isabelle Augenstein
Abstract: Retrieval-augmented generation (RAG) helps address the limitations of parametric knowledge embedded within a language model (LM). In real world settings, retrieved information can vary in complexity, yet most investigations of LM utilisation of context has been limited to synthetic text. We introduce DRUID (Dataset of Retrieved Unreliable, Insufficient and Difficult-to-understand contexts) with real-world queries and contexts manually annotated for stance. The dataset is based on the prototypical task of automated claim verification, for which automated retrieval of real-world evidence is crucial. We compare DRUID to synthetic datasets (CounterFact, ConflictQA) and find that artificial datasets often fail to represent the complexity and diversity of realistically retrieved context. We show that synthetic datasets exaggerate context characteristics rare in real retrieved data, which leads to inflated context utilisation results, as measured by our novel ACU score. Moreover, while previous work has mainly focused on singleton context characteristics to explain context utilisation, correlations between singleton context properties and ACU on DRUID are surprisingly small compared to other properties related to context source. Overall, our work underscores the need for real-world aligned context utilisation studies to represent and improve performance in real-world RAG settings.
Authors: Jaeheun Jung, Jaehyuk Lee, Changhae Jung, Hanyoung Kim, Bosung Jung, Donghun Lee
Abstract: Shock waves caused by earthquakes can be devastating. Generating realistic earthquake-caused ground motion waveforms help reducing losses in lives and properties, yet generative models for the task tend to generate subpar waveforms. We present High-fidelity Earthquake Groundmotion Generation System (HEGGS) and demonstrate its superior performance using earthquakes from North American, East Asian, and European regions. HEGGS exploits the intrinsic characteristics of earthquake dataset and learns the waveforms using an end-to-end differentiable generator containing conditional latent diffusion model and hi-fidelity waveform construction model. We show the learning efficiency of HEGGS by training it on a single GPU machine and validate its performance using earthquake databases from North America, East Asia, and Europe, using diverse criteria from waveform generation tasks and seismology. Once trained, HEGGS can generate three dimensional E-N-Z seismic waveforms with accurate P/S phase arrivals, envelope correlation, signal-to-noise ratio, GMPE analysis, frequency content analysis, and section plot analysis.
Authors: Soyeong Jeong, Kangsan Kim, Jinheon Baek, Sung Ju Hwang
Abstract: Retrieval-Augmented Generation (RAG) is a powerful strategy for improving the factual accuracy of models by retrieving external knowledge relevant to queries and incorporating it into the generation process. However, existing approaches primarily focus on text, with some recent advancements considering images, and they largely overlook videos, a rich source of multimodal knowledge capable of representing contextual details more effectively than any other modality. While very recent studies explore the use of videos in response generation, they either predefine query-associated videos without retrieval or convert videos into textual descriptions losing multimodal richness. To tackle these, we introduce VideoRAG, a framework that not only dynamically retrieves videos based on their relevance with queries but also utilizes both visual and textual information. The operation of VideoRAG is powered by recent Large Video Language Models (LVLMs), which enable the direct processing of video content to represent it for retrieval and the seamless integration of retrieved videos jointly with queries for response generation. Also, inspired by that the context size of LVLMs may not be sufficient to process all frames in extremely long videos and not all frames are equally important, we introduce a video frame selection mechanism to extract the most informative subset of frames, along with a strategy to extract textual information from videos (as it can aid the understanding of video content) when their subtitles are not available. We experimentally validate the effectiveness of VideoRAG, showcasing that it is superior to relevant baselines. Code is available at https://github.com/starsuzi/VideoRAG.
Authors: Yifan Zhang, Yifeng Liu, Huizhuo Yuan, Zhen Qin, Yang Yuan, Quanquan Gu, Andrew C Yao
Abstract: Scaling language models to handle longer input sequences typically necessitates large key-value (KV) caches, resulting in substantial memory overhead during inference. In this paper, we propose Tensor Product Attention (TPA), a novel attention mechanism that uses tensor decompositions to represent queries, keys, and values compactly, substantially shrinking the KV cache size at inference time. By factorizing these representations into contextual low-rank components and seamlessly integrating with Rotary Position Embedding (RoPE), TPA achieves improved model quality alongside memory efficiency. Based on TPA, we introduce the Tensor Product Attention Transformer,(T6), a new model architecture for sequence modeling. Through extensive empirical evaluation on language modeling tasks, we demonstrate that T6 surpasses or matches the performance of standard Transformer baselines, including Multi-Head Attention (MHA), Multi-Query Attention (MQA), Grouped-Query Attention (GQA), and Multi-Head Latent Attention (MLA) across various metrics, including perplexity and a range of established evaluation benchmarks. Notably, TPA's memory efficiency and computational efficiency at the decoding stage enable processing longer sequences under fixed resource constraints, addressing a critical scalability challenge in modern language models. The code is available at https://github.com/tensorgi/T6.
Authors: Chaoqi Wang, Zhuokai Zhao, Yibo Jiang, Zhaorun Chen, Chen Zhu, Yuxin Chen, Jiayi Liu, Lizhu Zhang, Xiangjun Fan, Hao Ma, Sinong Wang
Abstract: Recent advances in large language models (LLMs) have demonstrated significant progress in performing complex tasks. While Reinforcement Learning from Human Feedback (RLHF) has been effective in aligning LLMs with human preferences, it is susceptible to spurious correlations in reward modeling. Consequently, it often introduces biases-such as length bias, sycophancy, conceptual bias, and discrimination-that hinder the model's ability to capture true causal relationships. To address this, we propose a novel causal reward modeling approach that integrates causality to mitigate these spurious correlations. Our method enforces counterfactual invariance, ensuring reward predictions remain consistent when irrelevant variables are altered. Through experiments on both synthetic and real-world datasets, we show that our approach mitigates various types of spurious correlations effectively, resulting in more reliable and fair alignment of LLMs with human preferences. As a drop-in enhancement to the existing RLHF workflow, our causal reward modeling provides a practical way to improve the trustworthiness and fairness of LLM finetuning.
Authors: Rajvardhan Oak, Muhammad Haroon, Claire Jo, Magdalena Wojcieszak, Anshuman Chhabra
Abstract: Social media platforms utilize Machine Learning (ML) and Artificial Intelligence (AI) powered recommendation algorithms to maximize user engagement, which can result in inadvertent exposure to harmful content. Current moderation efforts, reliant on classifiers trained with extensive human-annotated data, struggle with scalability and adapting to new forms of harm. To address these challenges, we propose a novel re-ranking approach using Large Language Models (LLMs) in zero-shot and few-shot settings. Our method dynamically assesses and re-ranks content sequences, effectively mitigating harmful content exposure without requiring extensive labeled data. Alongside traditional ranking metrics, we also introduce two new metrics to evaluate the effectiveness of re-ranking in reducing exposure to harmful content. Through experiments on three datasets, three models and across three configurations, we demonstrate that our LLM-based approach significantly outperforms existing proprietary moderation approaches, offering a scalable and adaptable solution for harm mitigation.
Authors: Sangyeop Yeo, Seung-won Hwang, Yu-Seung Ma
Abstract: The use of Large Language Models (LLMs) for code generation has gained significant attention in recent years. Existing methods often aim to improve the quality of generated code by incorporating additional contextual information or guidance into input prompts. Many of these approaches adopt sequential reasoning strategies, mimicking human-like step-by-step thinking. However, such strategies may constrain flexibility, as they do not always align with the structured characteristics of programming languages. This paper introduces the Chain of Grounded Objectives (CGO), a method that embeds functional objectives into input prompts to enhance code generation. By leveraging appropriately structured objectives as input and avoiding explicit sequential procedures, CGO adapts effectively to the structured nature of programming tasks. Empirical evaluations demonstrate that CGO effectively enhances code generation, addressing limitations of existing approaches.
Authors: Shichang Zhang, Tessa Han, Usha Bhalla, Himabindu Lakkaraju
Abstract: The increasing complexity of AI systems has made understanding their behavior critical. Numerous interpretability methods have been developed to attribute model behavior to three key aspects: input features, training data, and internal model components, which emerged from explainable AI, data-centric AI, and mechanistic interpretability, respectively. However, these attribution methods are studied and applied rather independently, resulting in a fragmented landscape of methods and terminology. This position paper argues that feature, data, and component attribution methods share fundamental similarities, and a unified view of them benefits both interpretability and broader AI research. To this end, we first analyze popular methods for these three types of attributions and present a unified view demonstrating that these seemingly distinct methods employ similar techniques (such as perturbations, gradients, and linear approximations) over different aspects and thus differ primarily in their perspectives rather than techniques. Then, we demonstrate how this unified view enhances understanding of existing attribution methods, highlights shared concepts and evaluation criteria among these methods, and leads to new research directions both in interpretability research, by addressing common challenges and facilitating cross-attribution innovation, and in AI more broadly, with applications in model editing, steering, and regulation.
Authors: Haoran Luo, Haihong E, Yikai Guo, Qika Lin, Xiaobao Wu, Xinyu Mu, Wenhao Liu, Meina Song, Yifan Zhu, Luu Anh Tuan
Abstract: Knowledge Base Question Answering (KBQA) aims to answer natural language questions with a large-scale structured knowledge base (KB). Despite advancements with large language models (LLMs), KBQA still faces challenges in weak KB awareness, imbalance between effectiveness and efficiency, and high reliance on annotated data. To address these challenges, we propose KBQA-o1, a novel agentic KBQA method with Monte Carlo Tree Search (MCTS). It introduces a ReAct-based agent process for stepwise logical form generation with KB environment exploration. Moreover, it employs MCTS, a heuristic search method driven by policy and reward models, to balance agentic exploration's performance and search space. With heuristic exploration, KBQA-o1 generates high-quality annotations for further improvement by incremental fine-tuning. Experimental results show that KBQA-o1 outperforms previous low-resource KBQA methods with limited annotated data, boosting Llama-3.1-8B model's GrailQA F1 performance to 78.5% compared to 48.5% of the previous sota method with GPT-3.5-turbo. Our code is publicly available.
Authors: Wen Lai, Alexander Fraser, Ivan Titov
Abstract: Parameter-efficient fine-tuning (PEFT) methods, such as LoRA, are commonly used to adapt LLMs. However, the effectiveness of standard PEFT methods is limited in low-resource scenarios with only a few hundred examples. Recent advances in interpretability research have inspired the emergence of activation editing (or steering) techniques, which modify the activations of specific model components. Due to their extremely small parameter counts, these methods show promise for small datasets. However, their performance is highly dependent on identifying the correct modules to edit and often lacks stability across different datasets. In this paper, we propose Joint Localization and Activation Editing (JoLA), a method that jointly learns (1) which heads in the Transformer to edit (2) whether the intervention should be additive, multiplicative, or both and (3) the intervention parameters themselves - the vectors applied as additive offsets or multiplicative scalings to the head output. Through evaluations on three benchmarks spanning commonsense reasoning, natural language understanding, and natural language generation, we demonstrate that JoLA consistently outperforms existing methods. The code for the method is released at https://github.com/wenlai-lavine/jola.
Authors: Roman Tarasov, Petr Mokrov, Milena Gazdieva, Evgeny Burnaev, Alexander Korotin
Abstract: Neural network-based optimal transport (OT) is a recent and fruitful direction in the generative modeling community. It finds its applications in various fields such as domain translation, image super-resolution, computational biology and others. Among the existing OT approaches, of considerable interest are adversarial minimax solvers based on semi-dual formulations of OT problems. While promising, these methods lack theoretical investigation from a statistical learning perspective. Our work fills this gap by establishing upper bounds on the generalization error of an approximate OT map recovered by the minimax quadratic OT solver. Importantly, the bounds we derive depend solely on some standard statistical and mathematical properties of the considered functional classes (neural nets). While our analysis focuses on the quadratic OT, we believe that similar bounds could be derived for general OT case, paving the promising direction for future research.
Authors: Jiale Fu, Yuchu Jiang, Junkai Chen, Jiaming Fan, Xin Geng, Xu Yang
Abstract: Large Language Model (LLM) collaborative decoding techniques improve output quality by combining the outputs of multiple models at each generation step, but they incur high computational costs. In this paper, we introduce Collaborative decoding via Speculation (CoS), a novel framework that accelerates collaborative decoding without compromising performance. Inspired by Speculative Decoding--where a small proposal model generates tokens sequentially, and a larger target model verifies them in parallel, our approach builds on two key insights: (1) the verification distribution can be the combined distribution of both the proposal and target models, and (2) alternating each model as the proposer and verifier can further enhance efficiency. We generalize this method to collaboration among n models and theoretically prove that CoS is never slower than standard collaborative decoding, typically achieving faster speed. Extensive experiments demonstrate CoS is 1.11x-2.23x faster than standard collaborative decoding without compromising generation quality. Our code is available at https://github.com/Kamichanw/CoS/.
Authors: Zachary Cooper-Baldock, Stephen Turnock, Karl Sammut
Abstract: Autonomous Underwater Vehicles (AUVs) encounter significant energy, control and navigation challenges in complex underwater environments, particularly during close-proximity operations, such as launch and recovery (LAR), where fluid interactions and wake effects present additional navigational and energy challenges. Traditional path planning methods fail to incorporate these detailed wake structures, resulting in increased energy consumption, reduced control stability, and heightened safety risks. This paper presents a novel wake-informed, 3D path planning approach that fully integrates localized wake effects and global currents into the planning algorithm. Two variants of the A* algorithm - a current-informed planner and a wake-informed planner - are created to assess its validity and two neural network models are then trained to approximate these planners for real-time applications. Both the A* planners and NN models are evaluated using important metrics such as energy expenditure, path length, and encounters with high-velocity and turbulent regions. The results demonstrate a wake-informed A* planner consistently achieves the lowest energy expenditure and minimizes encounters with high-velocity regions, reducing energy consumption by up to 11.3%. The neural network models are observed to offer computational speedup of 6 orders of magnitude, but exhibit 4.51 - 19.79% higher energy expenditures and 9.81 - 24.38% less optimal paths. These findings underscore the importance of incorporating detailed wake structures into traditional path planning algorithms and the benefits of neural network approximations to enhance energy efficiency and operational safety for AUVs in complex 3D domains.
Authors: Alessio Russo, Aldo Pacchiano
Abstract: We study the policy evaluation problem in an online multi-reward multi-policy discounted setting, where multiple reward functions must be evaluated simultaneously for different policies. We adopt an $(\epsilon,\delta)$-PAC perspective to achieve $\epsilon$-accurate estimates with high confidence across finite or convex sets of rewards, a setting that has not been investigated in the literature. Building on prior work on Multi-Reward Best Policy Identification, we adapt the MR-NaS exploration scheme to jointly minimize sample complexity for evaluating different policies across different reward sets. Our approach leverages an instance-specific lower bound revealing how the sample complexity scales with a measure of value deviation, guiding the design of an efficient exploration policy. Although computing this bound entails a hard non-convex optimization, we propose an efficient convex approximation that holds for both finite and convex reward sets. Experiments in tabular domains demonstrate the effectiveness of this adaptive exploration scheme.
Authors: Adibvafa Fallahpour, Jun Ma, Alif Munim, Hongwei Lyu, Bo Wang
Abstract: Chest X-rays (CXRs) play an integral role in driving critical decisions in disease management and patient care. While recent innovations have led to specialized models for various CXR interpretation tasks, these solutions often operate in isolation, limiting their practical utility in clinical practice. We present MedRAX, the first versatile AI agent that seamlessly integrates state-of-the-art CXR analysis tools and multimodal large language models into a unified framework. MedRAX dynamically leverages these models to address complex medical queries without requiring additional training. To rigorously evaluate its capabilities, we introduce ChestAgentBench, a comprehensive benchmark containing 2,500 complex medical queries across 7 diverse categories. Our experiments demonstrate that MedRAX achieves state-of-the-art performance compared to both open-source and proprietary models, representing a significant step toward the practical deployment of automated CXR interpretation systems. Data and code have been publicly available at https://github.com/bowang-lab/MedRAX
Authors: Hongli Zhan, Muneeza Azmat, Raya Horesh, Junyi Jessy Li, Mikhail Yurochkin
Abstract: Aligning Large Language Models to integrate and reflect human values, especially for tasks that demand intricate human oversight, is arduous since it is resource-intensive and time-consuming to depend on human expertise for context-specific guidance. Prior work has utilized predefined sets of rules or principles to steer the behavior of models (Bai et al., 2022; Sun et al., 2023). However, these principles tend to be generic, making it challenging to adapt them to each individual input query or context. In this work, we present Situated-PRInciples (SPRI), a framework requiring minimal or no human effort that is designed to automatically generate guiding principles in real-time for each input query and utilize them to align each response. We evaluate SPRI on three tasks, and show that 1) SPRI can derive principles in a complex domain-specific task that leads to on-par performance as expert-crafted ones; 2) SPRI-generated principles lead to instance-specific rubrics that outperform prior LLM-as-a-judge frameworks; 3) using SPRI to generate synthetic SFT data leads to substantial improvement on truthfulness. We release our code and model generations at https://github.com/honglizhan/SPRI-public.
Authors: Daniel Beaglehole, Adityanarayanan Radhakrishnan, Enric Boix-Adser\`a, Mikhail Belkin
Abstract: Modern AI models contain much of human knowledge, yet understanding of their internal representation of this knowledge remains elusive. Characterizing the structure and properties of this representation will lead to improvements in model capabilities and development of effective safeguards. Building on recent advances in feature learning, we develop an effective, scalable approach for extracting linear representations of general concepts in large-scale AI models (language models, vision-language models, and reasoning models). We show how these representations enable model steering, through which we expose vulnerabilities, mitigate misaligned behaviors, and improve model capabilities. Additionally, we demonstrate that concept representations are remarkably transferable across human languages and combinable to enable multi-concept steering. Through quantitative analysis across hundreds of concepts, we find that newer, larger models are more steerable and steering can improve model capabilities beyond standard prompting. We show how concept representations are effective for monitoring misaligned content (hallucinations, toxic content). We demonstrate that predictive models built using concept representations are more accurate for monitoring misaligned content than using models that judge outputs directly. Together, our results illustrate the power of using internal representations to map the knowledge in AI models, advance AI safety, and improve model capabilities.
Authors: Yongchao Chen, Yilun Hao, Yueying Liu, Yang Zhang, Chuchu Fan
Abstract: Existing methods fail to effectively steer Large Language Models (LLMs) between textual reasoning and code generation, leaving symbolic computing capabilities underutilized. We introduce CodeSteer, an effective method for guiding LLM code/text generation. We construct a comprehensive benchmark SymBench comprising 37 symbolic tasks with adjustable complexity and also synthesize datasets of 12k multi-turn guidance/generation trajectories and 5.5k guidance comparison pairs. We fine-tune the Llama-3-8B model with a newly designed multi-turn supervised fine-tuning (SFT) and direct preference optimization (DPO). The resulting model, CodeSteerLLM, augmented with the proposed symbolic and self-answer checkers, effectively guides the code/text generation of larger models. Augmenting GPT-4o with CodeSteer raises its average performance score from 53.3 to 86.4, even outperforming the existing best LLM OpenAI o1 (82.7), o1-preview (74.8), and DeepSeek R1 (76.8) across all 37 tasks (28 seen, 9 unseen). Trained for GPT-4o, CodeSteer demonstrates superior generalizability, providing an average 41.8 performance boost on Claude, Mistral, and GPT-3.5. CodeSteer-guided LLMs fully harness symbolic computing to maintain strong performance on highly complex tasks. Models, Datasets, and Codes are available at https://github.com/yongchao98/CodeSteer-v1.0 and https://huggingface.co/yongchao98.
URLs: https://github.com/yongchao98/CodeSteer-v1.0, https://huggingface.co/yongchao98.
Authors: Elliot Meyerson, Xin Qiu
Abstract: Decomposing hard problems into subproblems often makes them easier and more efficient to solve. With large language models (LLMs) crossing critical reliability thresholds for a growing slate of capabilities, there is an increasing effort to decompose systems into sets of LLM-based agents, each of whom can be delegated sub-tasks. However, this decomposition (even when automated) is often intuitive, e.g., based on how a human might assign roles to members of a human team. How close are these role decompositions to optimal? This position paper argues that asymptotic analysis with LLM primitives is needed to reason about the efficiency of such decomposed systems, and that insights from such analysis will unlock opportunities for scaling them. By treating the LLM forward pass as the atomic unit of computational cost, one can separate out the (often opaque) inner workings of a particular LLM from the inherent efficiency of how a set of LLMs are orchestrated to solve hard problems. In other words, if we want to scale the deployment of LLMs to the limit, instead of anthropomorphizing LLMs, asymptotic analysis with LLM primitives should be used to reason about and develop more powerful decompositions of large problems into LLM agents.
Authors: Chen-Xiao Gao, Chenyang Wu, Mingjun Cao, Chenjun Xiao, Yang Yu, Zongzhang Zhang
Abstract: Behavior regularization, which constrains the policy to stay close to some behavior policy, is widely used in offline reinforcement learning (RL) to manage the risk of hazardous exploitation of unseen actions. Nevertheless, existing literature on behavior-regularized RL primarily focuses on explicit policy parameterizations, such as Gaussian policies. Consequently, it remains unclear how to extend this framework to more advanced policy parameterizations, such as diffusion models. In this paper, we introduce BDPO, a principled behavior-regularized RL framework tailored for diffusion-based policies, thereby combining the expressive power of diffusion policies and the robustness provided by regularization. The key ingredient of our method is to calculate the Kullback-Leibler (KL) regularization analytically as the accumulated discrepancies in reverse-time transition kernels along the diffusion trajectory. By integrating the regularization, we develop an efficient two-time-scale actor-critic RL algorithm that produces the optimal policy while respecting the behavior constraint. Comprehensive evaluations conducted on synthetic 2D tasks and continuous control tasks from the D4RL benchmark validate its effectiveness and superior performance.
Authors: Jeremy Kritz, Vaughn Robinson, Robert Vacareanu, Bijan Varjavand, Michael Choi, Bobby Gogov, Scale Red Team, Summer Yue, Willow E. Primack, Zifan Wang
Abstract: Large Language Models (LLMs) can be used to red team other models (e.g. jailbreaking) to elicit harmful contents. While prior works commonly employ open-weight models or private uncensored models for doing jailbreaking, as the refusal-training of strong LLMs (e.g. OpenAI o3) refuse to help jailbreaking, our work turn (almost) any black-box LLMs into attackers. The resulting $J_2$ (jailbreaking-to-jailbreak) attackers can effectively jailbreak the safeguard of target models using various strategies, both created by themselves or from expert human red teamers. In doing so, we show their strong but under-researched jailbreaking capabilities. Our experiments demonstrate that 1) prompts used to create $J_2$ attackers transfer across almost all black-box models; 2) an $J_2$ attacker can jailbreak a copy of itself, and this vulnerability develops rapidly over the past 12 months; 3) reasong models, such as Sonnet-3.7, are strong $J_2$ attackers compared to others. For example, when used against the safeguard of GPT-4o, $J_2$ (Sonnet-3.7) achieves 0.975 attack success rate (ASR), which matches expert human red teamers and surpasses the state-of-the-art algorithm-based attacks. Among $J_2$ attackers, $J_2$ (o3) achieves highest ASR (0.605) against Sonnet-3.5, one of the most robust models.
Authors: Zeli Su, Ziyin Zhang, Guixian Xu, Jianing Liu, XU Han, Ting Zhang, Yushuang Dong
Abstract: While multilingual language models like XLM-R have advanced multilingualism in NLP, they still perform poorly in extremely low-resource languages. This situation is exacerbated by the fact that modern LLMs such as LLaMA and Qwen support far fewer languages than XLM-R, making text generation models non-existent for many languages in the world. To tackle this challenge, we propose a novel framework for adapting multilingual encoders to text generation in extremely low-resource languages. By reusing the weights between the encoder and the decoder, our framework allows the model to leverage the learned semantic space of the encoder, enabling efficient learning and effective generalization in low-resource languages. Applying this framework to four Chinese minority languages, we present XLM-SWCM, and demonstrate its superior performance on various downstream tasks even when compared with much larger models.
Authors: Yi Wang, Fenghua Weng, Sibei Yang, Zhan Qin, Minlie Huang, Wenjie Wang
Abstract: Large Language Models (LLMs) are widely applied in decision making, but their deployment is threatened by jailbreak attacks, where adversarial users manipulate model behavior to bypass safety measures. Existing defense mechanisms, such as safety fine-tuning and model editing, either require extensive parameter modifications or lack precision, leading to performance degradation on general tasks, which is unsuitable to post-deployment safety alignment. To address these challenges, we propose DELMAN (Dynamic Editing for LLMs JAilbreak DefeNse), a novel approach leveraging direct model editing for precise, dynamic protection against jailbreak attacks. DELMAN directly updates a minimal set of relevant parameters to neutralize harmful behaviors while preserving the model's utility. To avoid triggering a safe response in benign context, we incorporate KL-divergence regularization to ensure the updated model remains consistent with the original model when processing benign queries. Experimental results demonstrate that DELMAN outperforms baseline methods in mitigating jailbreak attacks while preserving the model's utility, and adapts seamlessly to new attack instances, providing a practical and efficient solution for post-deployment model protection.
Authors: Jichan Chung, Irene Y. Chen
Abstract: The high cost of data labeling presents a major barrier to deploying machine learning systems at scale. Semi-supervised learning (SSL) mitigates this challenge by utilizing unlabeled data alongside limited labeled examples, while the emergence of foundation models (FMs) offers powerful zero-shot capabilities that can further reduce labeling cost. However, directly fine-tuning large FMs is often impractical in resource-constrained settings, and na\"ively using their pseudo-labels for unlabeled data can degrade performance due to its unreliablity or domain mismatch with target task. In this work, we introduce ZeroMatch, a novel SSL framework that integrates knowledge distillation with consistency-based learning to jointly leverage labeled data, unlabeled data, and pseudo-labels from FMs. ZeroMatch enables training compact student models using only FM inference, making it suitable for low-resource environments such as personal devices with limited compute. Experiments on six vision and language classification benchmarks show that ZeroMatch consistently outperforms standard SSL and zero-shot augmented methods, demonstrating its effectiveness and robustness across a range of foundation model qualities.
Authors: Sifan Zhou, Shuo Wang, Zhihang Yuan, Mingjia Shi, Yuzhang Shang, Dawei Yang
Abstract: Large Language Models (LLMs) fine-tuning technologies have achieved remarkable results. However, traditional LLM fine-tuning approaches face significant challenges: they require large Floating Point (FP) computation, raising privacy concerns when handling sensitive data, and are impractical for resource-constrained edge devices. While Parameter-Efficient Fine-Tuning (PEFT) techniques reduce trainable parameters, their reliance on floating-point arithmetic creates fundamental incompatibilities with edge hardware. In this work, we introduce a novel framework for on-device LLM fine-tuning that eliminates the need for floating-point operations in both inference and training, named GSQ-Tuning. At its core is the Group-Shared Exponents Integer format, which efficiently represents model parameters in integer format using shared exponents among parameter groups. When combined with LoRA-like adapters, this enables fully integer-based fine-tuning that is both memory and compute efficient. We demonstrate that our approach achieves accuracy comparable to BF16-based fine-tuning while significantly reducing 1.85x memory usage. Moreover, compared to FP8, our method can reduce 5x power consumption and 11x chip area with same performance, making large-scale model adaptation feasible on edge devices.
Authors: Hanlin Wang, Jian Wang, Chak Tou Leong, Wenjie Li
Abstract: Large language model (LLM)-based agents have shown promise in tackling complex tasks by interacting dynamically with the environment. Existing work primarily focuses on behavior cloning from expert demonstrations or preference learning through exploratory trajectory sampling. However, these methods often struggle to address long-horizon tasks, where suboptimal actions accumulate step by step, causing agents to deviate from correct task trajectories. To address this, we highlight the importance of timely calibration and the need to automatically construct calibration trajectories for training agents. We propose Step-Level Trajectory Calibration (STeCa), a novel framework for LLM agent learning. Specifically, STeCa identifies suboptimal actions through a step-level reward comparison during exploration. It constructs calibrated trajectories using LLM-driven reflection, enabling agents to learn from improved decision-making processes. We finally leverage these calibrated trajectories with successful trajectories for reinforced training. Extensive experiments demonstrate that STeCa significantly outperforms existing methods. Further analysis highlights that timely calibration enables agents to complete tasks with greater robustness. Our code and data are available at https://github.com/WangHanLinHenry/STeCa.
Authors: Pengcheng Huang, Zhenghao Liu, Yukun Yan, Haiyan Zhao, Xiaoyuan Yi, Hao Chen, Zhiyuan Liu, Maosong Sun, Tong Xiao, Ge Yu, Chenyan Xiong
Abstract: Large language models (LLMs) integrated with retrieval-augmented generation (RAG) have improved factuality by grounding outputs in external evidence. However, they remain susceptible to unfaithful generation, where outputs contradict retrieved context despite its relevance and accuracy. Existing approaches aiming to improve faithfulness primarily focus on enhancing the utilization of external context, but often overlook the persistent influence of internal parametric knowledge during generation. In this work, we investigate the internal mechanisms behind unfaithful generation and identify a subset of mid-to-deep feed-forward networks (FFNs) that are disproportionately activated in such cases. Building on this insight, we propose Parametric Knowledge Muting through FFN Suppression (ParamMute), a framework that improves contextual faithfulness by suppressing the activation of unfaithfulness-associated FFNs and calibrating the model toward retrieved knowledge. To evaluate our approach, we introduce CoFaithfulQA, a benchmark specifically designed to evaluate faithfulness in scenarios where internal knowledge conflicts with accurate external evidence. Experimental results show that ParamMute significantly enhances faithfulness across both CoFaithfulQA and the established ConFiQA benchmark, achieving substantial reductions in reliance on parametric memory. These findings underscore the importance of mitigating internal knowledge dominance and provide a new direction for improving LLM trustworthiness in RAG. All code will be released via GitHub.
Authors: Yanyang Li, Michael Lyu, Liwei Wang
Abstract: Solving complex tasks in a single attempt is challenging for large language models (LLMs). Iterative interaction with the environment and feedback is often required to achieve success, making effective feedback utilization a critical topic. Existing approaches either struggle with length generalization or rely on naive retries without leveraging prior information. In this paper, we introduce FTTT, a novel paradigm that formulates feedback utilization as an optimization problem at test time. Additionally, we propose a learnable test-time optimizer, OpTune, to effectively exploit feedback. Experiments on two LLMs across four reasoning datasets demonstrate that FTTT and OpTune achieve superior scalability and performance.
Authors: Zhipeng Cheng, Xiaoyu Xia, Hong Wang, Minghui Liwang, Ning Chen, Xuwei Fan, Xianbin Wang
Abstract: Edge inference (EI) has emerged as a promising paradigm to address the growing limitations of cloud-based Deep Neural Network (DNN) inference services, such as high response latency, limited scalability, and severe data privacy exposure. However, deploying DNN models on resource-constrained edge devices introduces additional challenges, including limited computation/storage resources, dynamic service demands, and heightened privacy risks. To tackle these issues, this paper presents a novel privacy-aware optimization framework that jointly addresses DNN model deployment, user-server association, and model partitioning, with the goal of minimizing long-term average inference delay under resource and privacy constraints. The problem is formulated as a complex, NP-hard stochastic optimization. To efficiently handle system dynamics and computational complexity, we employ a Lyapunov-based approach to transform the long-term objective into tractable per-slot decisions. Furthermore, we introduce a coalition formation game to enable adaptive user-server association and design a greedy algorithm for model deployment within each coalition. Extensive simulations demonstrate that the proposed algorithm significantly reduces inference delay and consistently satisfies privacy constraints, outperforming state-of-the-art baselines across diverse scenarios.
Authors: Kyungbok Lee, You Zhang, Zhiyao Duan
Abstract: The goal of Audio-Visual Segmentation (AVS) is to localize and segment the sounding source objects from video frames. Research on AVS suffers from data scarcity due to the high cost of fine-grained manual annotations. Recent works attempt to overcome the challenge of limited data by leveraging the vision foundation model, Segment Anything Model (SAM), prompting it with audio to enhance its ability to segment sounding source objects. While this approach alleviates the model's burden on understanding visual modality by utilizing knowledge of pre-trained SAM, it does not address the fundamental challenge of learning audio-visual correspondence with limited data. To address this limitation, we propose \textbf{AV2T-SAM}, a novel framework that bridges audio features with the text embedding space of pre-trained text-prompted SAM. Our method leverages multimodal correspondence learned from rich text-image paired datasets to enhance audio-visual alignment. Furthermore, we introduce a novel feature, $\mathbf{\textit{\textbf{f}}_{CLIP} \odot \textit{\textbf{f}}_{CLAP}}$, which emphasizes shared semantics of audio and visual modalities while filtering irrelevant noise. Our approach outperforms existing methods on the AVSBench dataset by effectively utilizing pre-trained segmentation models and cross-modal semantic alignment. The source code is released at https://github.com/bok-bok/AV2T-SAM.
Authors: Sijia Yao, Pengcheng Huang, Zhenghao Liu, Yu Gu, Yukun Yan, Shi Yu, Ge Yu
Abstract: Large language models (LLMs) have demonstrated significant potential in enhancing dense retrieval through query augmentation. However, most existing methods treat the LLM and the retriever as separate modules, overlooking the alignment between generation and ranking objectives. In this work, we propose ExpandR, a unified LLM-augmented dense retrieval framework that jointly optimizes both the LLM and the retriever. ExpandR employs the LLM to generate semantically rich query expansions, which are leveraged to enhance the retriever's training. Simultaneously, the LLM is trained using Direct Preference Optimization (DPO), guided by a carefully designed reward function that balances retrieval effectiveness and generation consistency. This joint optimization paradigm enables mutual adaptation between the LLM and the retriever, resulting in query expansions that are both informative and well-suited for retrieval. Experimental results on multiple benchmarks show that ExpandR consistently outperforms strong baselines, achieving more than a 5% improvement in retrieval performance. All codes are available at https://github.com/NEUIR/ExpandR.
Authors: Chaitanya Kapoor, Sudhanshu Srivastava, Meenakshi Khosla
Abstract: Understanding convergent learning -- the degree to which independently trained neural systems -- whether multiple artificial networks or brains and models -- arrive at similar internal representations -- is crucial for both neuroscience and AI. Yet, the literature remains narrow in scope -- typically examining just a handful of models with one dataset, relying on one alignment metric, and evaluating networks at a single post-training checkpoint. We present a large-scale audit of convergent learning, spanning dozens of vision models and thousands of layer-pair comparisons, to close these long-standing gaps. First, we pit three alignment families against one another -- linear regression (affine-invariant), orthogonal Procrustes (rotation-/reflection-invariant), and permutation/soft-matching (unit-order-invariant). We find that orthogonal transformations align representations nearly as effectively as more flexible linear ones, and although permutation scores are lower, they significantly exceed chance, indicating a privileged representational basis. Tracking convergence throughout training further shows that nearly all eventual alignment crystallizes within the first epoch -- well before accuracy plateaus -- indicating it is largely driven by shared input statistics and architectural biases, not by the final task solution. Finally, when models are challenged with a battery of out-of-distribution images, early layers remain tightly aligned, whereas deeper layers diverge in proportion to the distribution shift. These findings fill critical gaps in our understanding of representational convergence, with implications for neuroscience and AI.
Authors: Ruifeng Tan, Weixiang Hong, Jiayue Tang, Xibin Lu, Ruijun Ma, Xiang Zheng, Jia Li, Jiaqiang Huang, Tong-Yi Zhang
Abstract: Battery Life Prediction (BLP), which relies on time series data produced by battery degradation tests, is crucial for battery utilization, optimization, and production. Despite impressive advancements, this research area faces three key challenges. Firstly, the limited size of existing datasets impedes insights into modern battery life data. Secondly, most datasets are restricted to small-capacity lithium-ion batteries tested under a narrow range of diversity in labs, raising concerns about the generalizability of findings. Thirdly, inconsistent and limited benchmarks across studies obscure the effectiveness of baselines and leave it unclear if models popular in other time series fields are effective for BLP. To address these challenges, we propose BatteryLife, a comprehensive dataset and benchmark for BLP. BatteryLife integrates 16 datasets, offering a 2.5 times sample size compared to the previous largest dataset, and provides the most diverse battery life resource with batteries from 8 formats, 59 chemical systems, 9 operating temperatures, and 421 charge/discharge protocols, including both laboratory and industrial tests. Notably, BatteryLife is the first to release battery life datasets of zinc-ion batteries, sodium-ion batteries, and industry-tested large-capacity lithium-ion batteries. With the comprehensive dataset, we revisit the effectiveness of baselines popular in this and other time series fields. Furthermore, we propose CyclePatch, a plug-in technique that can be employed in various neural networks. Extensive benchmarking of 18 methods reveals that models popular in other time series fields can be unsuitable for BLP, and CyclePatch consistently improves model performance establishing state-of-the-art benchmarks. Moreover, BatteryLife evaluates model performance across aging conditions and domains. BatteryLife is available at https://github.com/Ruifeng-Tan/BatteryLife.
Authors: Thanh Le-Cong, Bach Le, Toby Murray
Abstract: Large Language Models (LLMs) are increasingly being used to automate programming tasks. Yet, LLMs' capabilities in reasoning about program semantics are still inadequately studied, leaving significant potential for further exploration. This paper introduces FormalBench, a comprehensive benchmark designed to evaluate LLMs' reasoning abilities on program semantics, particularly via the task of synthesizing formal program specifications to assist verifying program correctness. This task requires both comprehensive reasoning over all possible program executions and the generation of precise, syntactically correct expressions that adhere to formal syntax and semantics. Using this benchmark, we evaluated the ability of LLMs in synthesizing consistent and complete specifications. Our findings show that LLMs perform well with simple control flows but struggle with more complex structures, especially loops, even with advanced prompting. Additionally, LLMs exhibit limited robustness against semantic-preserving transformations. We also highlight common failure patterns and design self-repair prompts, improving success rates by 25%.
Authors: Jaa-Yeon Lee, Byunghee Cha, Jeongsol Kim, Jong Chul Ye
Abstract: While recent advancements in generative modeling have significantly improved text-image alignment, some residual misalignment between text and image representations still remains. Some approaches address this issue by fine-tuning models in terms of preference optimization, etc., which require tailored datasets. Orthogonal to these methods, we revisit the challenge from the perspective of representation alignment-an approach that has gained popularity with the success of REPresentation Alignment (REPA). We first argue that conventional text-to-image (T2I) diffusion models, typically trained on paired image and text data (i.e., positive pairs) by minimizing score matching or flow matching losses, is suboptimal from the standpoint of representation alignment. Instead, a better alignment can be achieved through contrastive learning that leverages existing dataset as both positive and negative pairs. To enable efficient alignment with pretrained models, we propose SoftREPA- a lightweight contrastive fine-tuning strategy that leverages soft text tokens for representation alignment. This approach improves alignment with minimal computational overhead by adding fewer than 1M trainable parameters to the pretrained model. Our theoretical analysis demonstrates that our method explicitly increases the mutual information between text and image representations, leading to enhanced semantic consistency. Experimental results across text-to-image generation and text-guided image editing tasks validate the effectiveness of our approach in improving the semantic consistency of T2I generative models.
Authors: Wenhao Hu, Jinhao Duan, Chunchen Wei, Li Zhang, Yue Zhang, Kaidi Xu
Abstract: The rapid advancement of large language models (LLMs) has significantly improved their performance in code generation tasks. However, existing code benchmarks remain static, consisting of fixed datasets with predefined problems. This makes them vulnerable to memorization during training, where LLMs recall specific test cases instead of generalizing to new problems, leading to data contamination and unreliable evaluation results. To address these issues, we introduce DynaCode, a dynamic, complexity-aware benchmark that overcomes the limitations of static datasets. DynaCode evaluates LLMs systematically using a complexity-aware metric, incorporating both code complexity and call-graph structures. DynaCode achieves large-scale diversity, generating up to 189 million unique nested code problems across four distinct levels of code complexity, referred to as units, and 16 types of call graphs. Results on 12 latest LLMs show an average performance drop of 16.8% to 45.7% compared to MBPP+, a static code generation benchmark, with performance progressively decreasing as complexity increases. This demonstrates DynaCode's ability to effectively differentiate LLMs. Additionally, by leveraging call graphs, we gain insights into LLM behavior, particularly their preference for handling subfunction interactions within nested code. Our benchmark and evaluation code are available at https://github.com/HWH-2000/DynaCode.
Authors: Shafiuddin Rehan Ahmed, Ankit Parag Shah, Quan Hung Tran, Vivek Khetan, Sukryool Kang, Ankit Mehta, Yujia Bao, Wei Wei
Abstract: Climate change has intensified the need for transparency and accountability in organizational practices, making Environmental, Social, and Governance (ESG) reporting increasingly crucial. Frameworks like the Global Reporting Initiative (GRI) and the new European Sustainability Reporting Standards (ESRS) aim to standardize ESG reporting, yet generating comprehensive reports remains challenging due to the considerable length of ESG documents and variability in company reporting styles. To facilitate ESG report automation, Retrieval-Augmented Generation (RAG) systems can be employed, but their development is hindered by a lack of labeled data suitable for training retrieval models. In this paper, we leverage an underutilized source of weak supervision -- the disclosure content index found in past ESG reports -- to create a comprehensive dataset, ESG-CID, for both GRI and ESRS standards. By extracting mappings between specific disclosure requirements and corresponding report sections, and refining them using a Large Language Model as a judge, we generate a robust training and evaluation set. We benchmark popular embedding models on this dataset and show that fine-tuning BERT-based models can outperform commercial embeddings and leading public models, even under temporal data splits for cross-report style transfer from GRI to ESRS. Data: https://huggingface.co/datasets/airefinery/esg_cid_retrieval
URLs: https://huggingface.co/datasets/airefinery/esg_cid_retrieval
Authors: Mingyang Song, Xiaoye Qu, Jiawei Zhou, Yu Cheng
Abstract: Large Vision-Language Models (LVLMs) have achieved significant progress in combining visual comprehension with language generation. Despite this success, the training data of LVLMs still suffers from Long-Tail (LT) problems, where the data distribution is highly imbalanced. Previous works have mainly focused on traditional VLM architectures, i.e., CLIP or ViT, and specific tasks such as recognition and classification. Nevertheless, the exploration of LVLM (e.g. LLaVA) and more general tasks (e.g. Visual Question Answering and Visual Reasoning) remains under-explored. In this paper, we first conduct an in-depth analysis of the LT issues in LVLMs and identify two core causes: the overrepresentation of head concepts and the underrepresentation of tail concepts. Based on the above observation, we propose an $\textbf{A}$daptive $\textbf{D}$ata $\textbf{R}$efinement Framework ($\textbf{ADR}$), which consists of two stages: $\textbf{D}$ata $\textbf{R}$ebalancing ($\textbf{DR}$) and $\textbf{D}$ata $\textbf{S}$ynthesis ($\textbf{DS}$). In the DR stage, we adaptively rebalance the redundant data based on entity distributions, while in the DS stage, we leverage Denoising Diffusion Probabilistic Models (DDPMs) and scarce images to supplement underrepresented portions. Through comprehensive evaluations across eleven benchmarks, our proposed ADR effectively mitigates the long-tail problem in the training data, improving the average performance of LLaVA 1.5 relatively by 4.36%, without increasing the training data volume.
Authors: Ricardo Bigolin Lanfredi, Yan Zhuang, Mark Finkelstein, Praveen Thoppey Srinivasan Balamuralikrishna, Luke Krembs, Brandon Khoury, Arthi Reddy, Pritam Mukherjee, Neil M. Rofsky, Ronald M. Summers
Abstract: Extracting structured labels from radiology reports has been employed to create vision models to simultaneously detect several types of abnormalities. However, existing works focus mainly on the chest region. Few works have been investigated on abdominal radiology reports due to more complex anatomy and a wider range of pathologies in the abdomen. We propose LEAVS (Large language model Extractor for Abdominal Vision Supervision). This labeler can annotate the certainty of presence and the urgency of seven types of abnormalities for nine abdominal organs on CT radiology reports. To ensure broad coverage, we chose abnormalities that encompass most of the finding types from CT reports. Our approach employs a specialized chain-of-thought prompting strategy for a locally-run LLM using sentence extraction and multiple-choice questions in a tree-based decision system. We demonstrate that the LLM can extract several abnormality types across abdominal organs with an average F1 score of 0.89, significantly outperforming competing labelers and humans. Additionally, we show that extraction of urgency labels achieved performance comparable to human annotations. Finally, we demonstrate that the abnormality labels contain valuable information for training a single vision model that classifies several organs as normal or abnormal. We release our code and structured annotations for a public CT dataset containing over 1,000 CT volumes.
Authors: Rochana Chaturvedi, Peyman Baghershahi, Sourav Medya, Barbara Di Eugenio
Abstract: Temporal information extraction from unstructured text is essential for contextualizing events and deriving actionable insights, particularly in the medical domain. We address the task of extracting clinical events and their temporal relations using the well-studied I2B2 2012 Temporal Relations Challenge corpus. This task is inherently challenging due to complex clinical language, long documents, and sparse annotations. We introduce GRAPHTREX, a novel method integrating span-based entity-relation extraction, clinical large pre-trained language models (LPLMs), and Heterogeneous Graph Transformers (HGT) to capture local and global dependencies. Our HGT component facilitates information propagation across the document through innovative global landmarks that bridge distant entities. Our method improves the state-of-the-art with 5.5% improvement in the tempeval $F_1$ score over the previous best and up to 8.9% improvement on long-range relations, which presents a formidable challenge. We further demonstrate generalizability by establishing a strong baseline on the E3C corpus. This work not only advances temporal information extraction but also lays the groundwork for improved diagnostic and prognostic models through enhanced temporal reasoning.
Authors: Christoforos N. Spartalis, Theodoros Semertzidis, Efstratios Gavves, Petros Daras
Abstract: We present LoTUS, a novel Machine Unlearning (MU) method that eliminates the influence of training samples from pre-trained models, avoiding retraining from scratch. LoTUS smooths the prediction probabilities of the model up to an information-theoretic bound, mitigating its over-confidence stemming from data memorization. We evaluate LoTUS on Transformer and ResNet18 models against eight baselines across five public datasets. Beyond established MU benchmarks, we evaluate unlearning on ImageNet1k, a large-scale dataset, where retraining is impractical, simulating real-world conditions. Moreover, we introduce the novel Retrain-Free Jensen-Shannon Divergence (RF-JSD) metric to enable evaluation under real-world conditions. The experimental results show that LoTUS outperforms state-of-the-art methods in terms of both efficiency and effectiveness. Code: https://github.com/cspartalis/LoTUS.
Authors: Yubo Li, Yidi Miao, Xueying Ding, Ramayya Krishnan, Rema Padman
Abstract: Large Language Models (LLMs) have shown remarkable capabilities across various tasks, but their deployment in high-stake domains requires consistent performance across multiple interaction rounds. This paper introduces a comprehensive framework for evaluating and improving LLM response consistency, making three key contributions. First, we propose a novel Position-Weighted Consistency (PWC) score that captures both the importance of early-stage stability and recovery patterns in multi-turn interactions. Second, we present a carefully curated benchmark dataset spanning diverse domains and difficulty levels, specifically designed to evaluate LLM consistency under various challenging follow-up scenarios. Third, we introduce Confidence-Aware Response Generation (CARG), a framework that significantly improves response stability by incorporating model confidence signals into the generation process. Empirical results demonstrate that CARG significantly improves response stability without sacrificing accuracy, underscoring its potential for reliable LLM deployment in critical applications.
Authors: Ziwei Zhang, Juan Wen, Wanli Peng, Zhengxian Wu, Yinghan Zhou, Yiming Xue
Abstract: In-Context Learning (ICL) and efficient fine-tuning methods significantly enhanced the efficiency of applying Large Language Models (LLMs) to downstream tasks. However, they also raise concerns about the imitation and infringement of personal creative data. Current methods for data copyright protection primarily focuses on content security but lacks effectiveness in protecting the copyrights of text styles. In this paper, we introduce a novel implicit zero-watermarking scheme, namely MiZero. This scheme establishes a precise watermark domain to protect the copyrighted style, surpassing traditional watermarking methods that distort the style characteristics. Specifically, we employ LLMs to extract condensed-lists utilizing the designed instance delimitation mechanism. These lists guide MiZero in generating the watermark. Extensive experiments demonstrate that MiZero effectively verifies text style copyright ownership against AI imitation.
Authors: Olawale Salaudeen, Nicole Chiou, Shiny Weng, Sanmi Koyejo
Abstract: Spurious correlations are unstable statistical associations that hinder robust decision-making. Conventional wisdom suggests that models relying on such correlations will fail to generalize out-of-distribution (OOD), especially under strong distribution shifts. However, empirical evidence challenges this view as naive in-distribution empirical risk minimizers often achieve the best OOD accuracy across popular OOD generalization benchmarks. In light of these results, we propose a different perspective: many widely used benchmarks for evaluating robustness to spurious correlations are misspecified. Specifically, they fail to include shifts in spurious correlations that meaningfully impact OOD generalization, making them unsuitable for evaluating the benefit of removing such correlations. We establish conditions under which a distribution shift can reliably assess a model's reliance on spurious correlations. Crucially, under these conditions, we should not observe a strong positive correlation between in-distribution and OOD accuracy, often called "accuracy on the line." Yet, most state-of-the-art benchmarks exhibit this pattern, suggesting they do not effectively assess robustness. Our findings expose a key limitation in current benchmarks used to evaluate domain generalization algorithms, that is, models designed to avoid spurious correlations. We highlight the need to rethink how robustness to spurious correlations is assessed, identify well-specified benchmarks the field should prioritize, and enumerate strategies for designing future benchmarks that meaningfully reflect robustness under distribution shift.
Authors: Shuofei Qiao, Zhisong Qiu, Baochang Ren, Xiaobin Wang, Xiangyuan Ru, Ningyu Zhang, Xiang Chen, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen
Abstract: Large Language Models (LLMs) have achieved considerable performance across various agentic planning tasks. However, traditional agent planning approaches adopt a "flood irrigation" methodology that indiscriminately injects gold trajectories, external feedback, and domain knowledge into agent models. This practice overlooks the fundamental human cognitive principle of situational self-awareness during decision-making-the ability to dynamically assess situational demands and strategically employ resources during decision-making. We propose agentic knowledgeable self-awareness to address this gap, a novel paradigm enabling LLM-based agents to autonomously regulate knowledge utilization. Specifically, we propose KnowSelf, a data-centric approach that applies agents with knowledgeable self-awareness like humans. Concretely, we devise a heuristic situation judgement criterion to mark special tokens on the agent's self-explored trajectories for collecting training data. Through a two-stage training process, the agent model can switch between different situations by generating specific special tokens, achieving optimal planning effects with minimal costs. Our experiments demonstrate that KnowSelf can outperform various strong baselines on different tasks and models with minimal use of external knowledge. Code is available at https://github.com/zjunlp/KnowSelf.
Authors: Aikaterini Maria Panteleaki, Konstantinos Balaskas, Georgios Zervakis, Hussam Amrouch, Iraklis Anagnostopoulos
Abstract: As Deep Neural Networks (DNNs) continue to drive advancements in artificial intelligence, the design of hardware accelerators faces growing concerns over embodied carbon footprint due to complex fabrication processes. 3D integration improves performance but introduces sustainability challenges, making carbon-aware optimization essential. In this work, we propose a carbon-efficient design methodology for 3D DNN accelerators, leveraging approximate computing and genetic algorithm-based design space exploration to optimize Carbon Delay Product (CDP). By integrating area-efficient approximate multipliers into Multiply-Accumulate (MAC) units, our approach effectively reduces silicon area and fabrication overhead while maintaining high computational accuracy. Experimental evaluations across three technology nodes (45nm, 14nm, and 7nm) show that our method reduces embodied carbon by up to 30% with negligible accuracy drop.
Authors: Mete Erdogan, Cengiz Pehlevan, Alper T. Erdogan
Abstract: We introduce Error Broadcast and Decorrelation (EBD), a novel learning framework for neural networks that addresses credit assignment by directly broadcasting output errors to individual layers, circumventing weight transport of backpropagation. EBD is rigorously grounded in the stochastic orthogonality property of Minimum Mean Square Error estimators. This fundamental principle states that the error of an optimal estimator is orthogonal to functions of the input. Guided by this insight, EBD defines layerwise loss functions that directly penalize correlations between layer activations and output errors, thereby establishing a principled foundation for error broadcasting. This theoretically sound mechanism naturally leads to the experimentally observed three-factor learning rule and integrates with biologically plausible frameworks to enhance performance and plausibility. Numerical experiments demonstrate EBD's competitive or better performance against other error-broadcast methods on benchmark datasets. Our findings establish EBD as an efficient, biologically plausible, and principled alternative for neural network training.
Authors: Sahil Rajesh Dhayalkar
Abstract: We propose a combinatorial and graph-theoretic theory of dropout by modeling training as a random walk over a high-dimensional graph of binary subnetworks. Each node represents a masked version of the network, and dropout induces stochastic traversal across this space. We define a subnetwork contribution score that quantifies generalization and show that it varies smoothly over the graph. Using tools from spectral graph theory, PAC-Bayes analysis, and combinatorics, we prove that generalizing subnetworks form large, connected, low-resistance clusters, and that their number grows exponentially with network width. This reveals dropout as a mechanism for sampling from a robust, structured ensemble of well-generalizing subnetworks with built-in redundancy. Extensive experiments validate every theoretical claim across diverse architectures. Together, our results offer a unified foundation for understanding dropout and suggest new directions for mask-guided regularization and subnetwork optimization.
Authors: Jianhao Yan, Yafu Li, Zican Hu, Zhi Wang, Ganqu Cui, Xiaoye Qu, Yu Cheng, Yue Zhang
Abstract: Recent advances in large reasoning models (LRMs) demonstrate that sophisticated behaviors such as multi-step reasoning and self-reflection can emerge via reinforcement learning with verifiable rewards~(\textit{RLVR}). However, existing \textit{RLVR} approaches are inherently ``on-policy'', limiting learning to a model's own outputs and failing to acquire reasoning abilities beyond its initial capabilities. To address this issue, we introduce \textbf{LUFFY} (\textbf{L}earning to reason \textbf{U}nder o\textbf{FF}-polic\textbf{Y} guidance), a framework that augments \textit{RLVR} with off-policy reasoning traces. LUFFY dynamically balances imitation and exploration by combining off-policy demonstrations with on-policy rollouts during training. Specifically, LUFFY combines the Mixed-Policy GRPO framework, which has a theoretically guaranteed convergence rate, alongside policy shaping via regularized importance sampling to avoid superficial and rigid imitation during mixed-policy training. Compared with previous RLVR methods, LUFFY achieves an over \textbf{+6.4} average gain across six math benchmarks and an advantage of over \textbf{+6.2} points in out-of-distribution tasks. Most significantly, we show that LUFFY successfully trains weak models in scenarios where on-policy RLVR completely fails. These results provide compelling evidence that LUFFY transcends the fundamental limitations of on-policy RLVR and demonstrates the great potential of utilizing off-policy guidance in RLVR.
Authors: Tengchao Zhang, Yonglin Tian, Fei Lin, Jun Huang, Patrik P. S\"uli, Rui Qin, Fei-Yue Wang
Abstract: With the increasing demand for heterogeneous Unmanned Aerial Vehicle (UAV) swarms to perform complex tasks in urban environments, system design now faces major challenges, including efficient semantic understanding, flexible task planning, and the ability to dynamically adjust coordination strategies in response to evolving environmental conditions and continuously changing task requirements. To address the limitations of existing approaches, this paper proposes coordination field agentic system for coordinating heterogeneous UAV swarms in complex urban scenarios. In this system, large language models (LLMs) is responsible for interpreting high-level human instructions and converting them into executable commands for the UAV swarms, such as patrol and target tracking. Subsequently, a Coordination field mechanism is proposed to guide UAV motion and task selection, enabling decentralized and adaptive allocation of emergent tasks. A total of 50 rounds of comparative testing were conducted across different models in a 2D simulation space to evaluate their performance. Experimental results demonstrate that the proposed system achieves superior performance in terms of task coverage, response time, and adaptability to dynamic changes.
Authors: Xinyue Zeng, Haohui Wang, Junhong Lin, Jun Wu, Tyler Cody, Dawei Zhou
Abstract: The proliferation of open-sourced Large Language Models (LLMs) and diverse downstream tasks necessitates efficient model selection, given the impracticality of fine-tuning all candidates due to computational constraints. Despite the recent advances in LLM selection, a fundamental research question largely remains nascent: how can we model the dynamic behaviors of LLMs during fine-tuning, thereby enhancing our understanding of their generalization performance across diverse downstream tasks? In this work, we propose a novel theoretical framework that provides a proper lens to assess the generalization capabilities of LLMs, thereby enabling accurate and efficient LLM selection for downstream applications. In particular, we first derive a PAC-Bayesian Generalization Bound that unveils fine-tuning dynamics of LLMs and then introduce LENSLLM, a Neural Tangent Kernel (NTK)-based Rectified Scaling Model that enables accurate performance predictions across diverse tasks while maintaining computational efficiency. Extensive empirical results on 3 large-scale benchmarks demonstrate that our model achieves up to 91.1% accuracy and reduces up to 88.5% computational cost in LLM selection, outperforming 5 state-of-the-art methods. We open-source our proposed LENSLLM model and corresponding results at LensLLM.io.
Authors: Soumik Dey, Hansi Wu, Binbin Li
Abstract: E-commerce sellers are recommended keyphrases based on their inventory on which they advertise to increase buyer engagement (clicks/sales). The relevance of advertiser keyphrases plays an important role in preventing the inundation of search systems with numerous irrelevant items that compete for attention in auctions, in addition to maintaining a healthy seller perception. In this work, we describe the shortcomings of training Advertiser keyphrase relevance filter models on click/sales/search relevance signals and the importance of aligning with human judgment, as sellers have the power to adopt or reject said keyphrase recommendations. In this study, we frame Advertiser keyphrase relevance as a complex interaction between 3 dynamical systems -- seller judgment, which influences seller adoption of our product, Advertising, which provides the keyphrases to bid on, and Search, who holds the auctions for the same keyphrases. This study discusses the practicalities of using human judgment via a case study at eBay Advertising and demonstrate that using LLM-as-a-judge en-masse as a scalable proxy for seller judgment to train our relevance models achieves a better harmony across the three systems -- provided that they are bound by a meticulous evaluation framework grounded in business metrics.
Authors: Dario Garcia-Gasulla, Jordi Bayarri-Planas, Ashwin Kumar Gururajan, Enrique Lopez-Cuena, Adrian Tormos, Daniel Hinjos, Pablo Bernabeu-Perez, Anna Arias-Duart, Pablo Agustin Martin-Torres, Marta Gonzalez-Mallo, Sergio Alvarez-Napagao, Eduard Ayguad\'e-Parra, Ulises Cort\'es
Abstract: Purpose: With advancements in Large Language Models (LLMs) for healthcare, the need arises for competitive open-source models to protect the public interest. This work contributes to the field of open medical LLMs by optimizing key stages of data preprocessing and training, while showing how to improve model safety (through DPO) and efficacy (through RAG). The evaluation methodology used, which includes four different types of tests, defines a new standard for the field. The resultant models, shown to be competitive with the best private alternatives, are released with a permisive license. Methods: Building on top of strong base models like Llama 3.1 and Qwen 2.5, Aloe Beta uses a custom dataset to enhance public data with synthetic Chain of Thought examples. The models undergo alignment with Direct Preference Optimization, emphasizing ethical and policy-aligned performance in the presence of jailbreaking attacks. Evaluation includes close-ended, open-ended, safety and human assessments, to maximize the reliability of results. Results: Recommendations are made across the entire pipeline, backed by the solid performance of the Aloe Family. These models deliver competitive performance across healthcare benchmarks and medical fields, and are often preferred by healthcare professionals. On bias and toxicity, the Aloe Beta models significantly improve safety, showing resilience to unseen jailbreaking attacks. For a responsible release, a detailed risk assessment specific to healthcare is attached to the Aloe Family models. Conclusion: The Aloe Beta models, and the recipe that leads to them, are a significant contribution to the open-source medical LLM field, offering top-of-the-line performance while maintaining high ethical requirements. This work sets a new standard for developing and reporting aligned LLMs in healthcare.
Authors: Yuqin Lan, Laurence T. Yang
Abstract: In the era of rapid development of social media, social recommendation systems as hybrid recommendation systems have been widely applied. Existing methods capture interest similarity between users to filter out interest-irrelevant relations in social networks that inevitably decrease recommendation accuracy, however, limited research has a focus on the mutual influence of semantic information between the social network and the user-item interaction network for further improving social recommendation. To address these issues, we introduce a social \underline{r}ecommendation model with ro\underline{bu}st g\underline{r}aph denoisin\underline{g}-augmentation fusion and multi-s\underline{e}mantic Modeling(Burger). Specifically, we firstly propose to construct a social tensor in order to smooth the training process of the model. Then, a graph convolutional network and a tensor convolutional network are employed to capture user's item preference and social preference, respectively. Considering the different semantic information in the user-item interaction network and the social network, a bi-semantic coordination loss is proposed to model the mutual influence of semantic information. To alleviate the interference of interest-irrelevant relations on multi-semantic modeling, we further use Bayesian posterior probability to mine potential social relations to replace social noise. Finally, the sliding window mechanism is utilized to update the social tensor as the input for the next iteration. Extensive experiments on three real datasets show Burger has a superior performance compared with the state-of-the-art models.
Authors: Steven Song, Morgan Borjigin-Wang, Irene Madejski, Robert L. Grossman
Abstract: The Cancer Genome Atlas (TCGA) has enabled novel discoveries and served as a large-scale reference through its harmonized genomics, clinical, and image data. Prior studies have trained bespoke cancer survival prediction models from unimodal or multimodal TCGA data. A modern paradigm in biomedical deep learning is the development of foundation models (FMs) to derive meaningful feature embeddings, agnostic to a specific modeling task. Biomedical text especially has seen growing development of FMs. While TCGA contains free-text data as pathology reports, these have been historically underutilized. Here, we investigate the feasibility of training classical, multimodal survival models over zero-shot embeddings extracted by FMs. We show the ease and additive effect of multimodal fusion, outperforming unimodal models. We demonstrate the benefit of including pathology report text and rigorously evaluate the effect of model-based text summarization and hallucination. Overall, we modernize survival modeling by leveraging FMs and information extraction from pathology reports.
Authors: Ruilin Liu, Zhixiao Zhao, Jieqiong Li, Chang Liu, Dongbo Wang
Abstract: The rapid development of large language models (LLMs) has provided significant support and opportunities for the advancement of domain-specific LLMs. However, fine-tuning these large models using Intangible Cultural Heritage (ICH) data inevitably faces challenges such as bias, incorrect knowledge inheritance, and catastrophic forgetting. To address these issues, we propose a novel training method that integrates a bidirectional chains of thought and a reward mechanism. This method is built upon ICH-Qwen, a large language model specifically designed for the field of intangible cultural heritage. The proposed method enables the model to not only perform forward reasoning but also enhances the accuracy of the generated answers by utilizing reverse questioning and reverse reasoning to activate the model's latent knowledge. Additionally, a reward mechanism is introduced during training to optimize the decision-making process. This mechanism improves the quality of the model's outputs through structural and content evaluations with different weighting schemes. We conduct comparative experiments on ICH-Qwen, with results demonstrating that our method outperforms 0-shot, step-by-step reasoning, knowledge distillation, and question augmentation methods in terms of accuracy, Bleu-4, and Rouge-L scores on the question-answering task. Furthermore, the paper highlights the effectiveness of combining the bidirectional chains of thought and reward mechanism through ablation experiments. In addition, a series of generalizability experiments are conducted, with results showing that the proposed method yields improvements on various domain-specific datasets and advanced models in areas such as Finance, Wikidata, and StrategyQA. This demonstrates that the method is adaptable to multiple domains and provides a valuable approach for model training in future applications across diverse fields.
Authors: Fujun Zhang, Xiaoying Fan, XiangDong Su, Guanglai Gao
Abstract: Fine-tuning pre-trained language models (PLMs) has become a dominant paradigm in applying PLMs to downstream tasks. However, with limited fine-tuning, PLMs still struggle with the discrepancies between the representation obtained from the PLMs' encoder and the optimal input to the PLMs' decoder. This paper tackles this challenge by learning to calibrate the representation of PLMs in the latent space. In the proposed representation calibration method (RepCali), we integrate a specific calibration block to the latent space after the encoder and use the calibrated output as the decoder input. The merits of the proposed RepCali include its universality to all PLMs with encoder-decoder architectures, its plug-and-play nature, and ease of implementation. Extensive experiments on 25 PLM-based models across 8 tasks (including both English and Chinese datasets) demonstrate that the proposed RepCali offers desirable enhancements to PLMs (including LLMs) and significantly improves the performance of downstream tasks. Comparison experiments across 4 benchmark tasks indicate that RepCali is superior to the representative fine-tuning baselines.
Authors: Sahil Rajesh Dhayalkar
Abstract: We develop a novel theoretical framework for analyzing ReLU neural networks through the lens of a combinatorial object we term the ReLU Transition Graph (RTG). In this graph, each node corresponds to a linear region induced by the network's activation patterns, and edges connect regions that differ by a single neuron flip. Building on this structure, we derive a suite of new theoretical results connecting RTG geometry to expressivity, generalization, and robustness. Our contributions include tight combinatorial bounds on RTG size and diameter, a proof of RTG connectivity, and graph-theoretic interpretations of VC-dimension. We also relate entropy and average degree of the RTG to generalization error. Each theoretical result is rigorously validated via carefully controlled experiments across varied network depths, widths, and data regimes. This work provides the first unified treatment of ReLU network structure via graph theory and opens new avenues for compression, regularization, and complexity control rooted in RTG analysis.
Authors: Sahil Rajesh Dhayalkar
Abstract: We present a complete theoretical and empirical framework establishing feedforward neural networks as universal finite-state machines (N-FSMs). Our results prove that finite-depth ReLU and threshold networks can exactly simulate deterministic finite automata (DFAs) by unrolling state transitions into depth-wise neural layers, with formal characterizations of required depth, width, and state compression. We demonstrate that DFA transitions are linearly separable, binary threshold activations allow exponential compression, and Myhill-Nerode equivalence classes can be embedded into continuous latent spaces while preserving separability. We also formalize the expressivity boundary: fixed-depth feedforward networks cannot recognize non-regular languages requiring unbounded memory. Unlike prior heuristic or probing-based studies, we provide constructive proofs and design explicit DFA-unrolled neural architectures that empirically validate every claim. Our results bridge deep learning, automata theory, and neural-symbolic computation, offering a rigorous blueprint for how discrete symbolic processes can be realized in continuous neural systems.
Authors: Yu Fan, Jingwei Ni, Jakob Merane, Etienne Salimbeni, Yang Tian, Yoan Hermstr\"uwer, Yinya Huang, Mubashara Akhtar, Florian Geering, Oliver Dreyer, Daniel Brunner, Markus Leippold, Mrinmaya Sachan, Alexander Stremitzer, Christoph Engel, Elliott Ash, Joel Niklaus
Abstract: Long-form legal reasoning remains a key challenge for large language models (LLMs) in spite of recent advances in test-time scaling. We introduce LEXam, a novel benchmark derived from 340 law exams spanning 116 law school courses across a range of subjects and degree levels. The dataset comprises 4,886 law exam questions in English and German, including 2,841 long-form, open-ended questions and 2,045 multiple-choice questions. Besides reference answers, the open questions are also accompanied by explicit guidance outlining the expected legal reasoning approach such as issue spotting, rule recall, or rule application. Our evaluation on both open-ended and multiple-choice questions present significant challenges for current LLMs; in particular, they notably struggle with open questions that require structured, multi-step legal reasoning. Moreover, our results underscore the effectiveness of the dataset in differentiating between models with varying capabilities. Adopting an LLM-as-a-Judge paradigm with rigorous human expert validation, we demonstrate how model-generated reasoning steps can be evaluated consistently and accurately. Our evaluation setup provides a scalable method to assess legal reasoning quality beyond simple accuracy metrics. Project page: https://lexam-benchmark.github.io/
Authors: Von Ralph Dane Marquez Herbuela, Yukie Nagai
Abstract: Understanding how humans express and synchronize emotions across multiple communication channels particularly facial expressions and speech has significant implications for emotion recognition systems and human computer interaction. Motivated by the notion that non-overlapping speech promotes clearer emotional coordination, while overlapping speech disrupts synchrony, this study examines how these conversational dynamics shape the spatial and temporal alignment of arousal and valence across facial and vocal modalities. Using dyadic interactions from the IEMOCAP dataset, we extracted continuous emotion estimates via EmoNet (facial video) and a Wav2Vec2-based model (speech audio). Segments were categorized based on speech overlap, and emotional alignment was assessed using Pearson correlation, lag adjusted analysis, and Dynamic Time Warping (DTW). Across analyses, non overlapping speech was associated with more stable and predictable emotional synchrony than overlapping speech. While zero-lag correlations were low and not statistically different, non overlapping speech showed reduced variability, especially for arousal. Lag adjusted correlations and best-lag distributions revealed clearer, more consistent temporal alignment in these segments. In contrast, overlapping speech exhibited higher variability and flatter lag profiles, though DTW indicated unexpectedly tighter alignment suggesting distinct coordination strategies. Notably, directionality patterns showed that facial expressions more often preceded speech during turn-taking, while speech led during simultaneous vocalizations. These findings underscore the importance of conversational structure in regulating emotional communication and provide new insight into the spatial and temporal dynamics of multimodal affective alignment in real world interaction.
Authors: Jihwan Lee, Kevin Huang, Kleanthis Avramidis, Simon Pistrosch, Monica Gonzalez-Machorro, Yoonjeong Lee, Bj\"orn Schuller, Louis Goldstein, Shrikanth Narayanan
Abstract: We present a model for predicting articulatory features from surface electromyography (EMG) signals during speech production. The proposed model integrates convolutional layers and a Transformer block, followed by separate predictors for articulatory features. Our approach achieves a high prediction correlation of approximately 0.9 for most articulatory features. Furthermore, we demonstrate that these predicted articulatory features can be decoded into intelligible speech waveforms. To our knowledge, this is the first method to decode speech waveforms from surface EMG via articulatory features, offering a novel approach to EMG-based speech synthesis. Additionally, we analyze the relationship between EMG electrode placement and articulatory feature predictability, providing knowledge-driven insights for optimizing EMG electrode configurations. The source code and decoded speech samples are publicly available.
Authors: Yakun Zhu, Zhongzhen Huang, Linjie Mu, Yutong Huang, Wei Nie, Jiaji Liu, Shaoting Zhang, Pengfei Liu, Xiaofan Zhang
Abstract: The emergence of groundbreaking large language models capable of performing complex reasoning tasks holds significant promise for addressing various scientific challenges, including those arising in complex clinical scenarios. To enable their safe and effective deployment in real-world healthcare settings, it is urgently necessary to benchmark the diagnostic capabilities of current models systematically. Given the limitations of existing medical benchmarks in evaluating advanced diagnostic reasoning, we present DiagnosisArena, a comprehensive and challenging benchmark designed to rigorously assess professional-level diagnostic competence. DiagnosisArena consists of 1,113 pairs of segmented patient cases and corresponding diagnoses, spanning 28 medical specialties, deriving from clinical case reports published in 10 top-tier medical journals. The benchmark is developed through a meticulous construction pipeline, involving multiple rounds of screening and review by both AI systems and human experts, with thorough checks conducted to prevent data leakage. Our study reveals that even the most advanced reasoning models, o3, o1, and DeepSeek-R1, achieve only 51.12%, 31.09%, and 17.79% accuracy, respectively. This finding highlights a significant generalization bottleneck in current large language models when faced with clinical diagnostic reasoning challenges. Through DiagnosisArena, we aim to drive further advancements in AI's diagnostic reasoning capabilities, enabling more effective solutions for real-world clinical diagnostic challenges. We provide the benchmark and evaluation tools for further research and development https://github.com/SPIRAL-MED/DiagnosisArena.
Authors: Jennifer D'Souza, Hamed Babaei Giglou, Quentin M\"unch
Abstract: Large Language Models (LLMs) drive scientific question-answering on modern search engines, yet their evaluation robustness remains underexplored. We introduce YESciEval, an open-source framework that combines fine-grained rubric-based assessment with reinforcement learning to mitigate optimism bias in LLM evaluators. We release multidisciplinary scienceQ&A datasets, including adversarial variants, with evaluation scores from multiple LLMs. Independent of proprietary models and human feedback, our approach enables scalable, cost-free evaluation. By advancing reliable LLM-as-a-judge models, this work supports AI alignment and fosters robust, transparent evaluation essential for scientific inquiry.
Authors: Roberto Morabito, SiYoung Jang
Abstract: The rapid adoption of generative AI (GenAI), particularly Large Language Models (LLMs), has exposed critical limitations of cloud-centric deployments, including latency, cost, and privacy concerns. Meanwhile, Small Language Models (SLMs) are emerging as viable alternatives for resource-constrained edge environments, though they often lack the capabilities of their larger counterparts. This article explores the potential of collaborative inference systems that leverage both edge and cloud resources to address these challenges. By presenting distinct cooperation strategies alongside practical design principles and experimental insights, we offer actionable guidance for deploying GenAI across the computing continuum.
Authors: SiYoung Jang, Roberto Morabito
Abstract: The widespread adoption of Language Models (LMs) across industries is driving interest in deploying these services across the computing continuum, from the cloud to the network edge. This shift aims to reduce costs, lower latency, and improve reliability and privacy. Small Language Models (SLMs), enabled by advances in model compression, are central to this shift, offering a path to on-device inference on resource-constrained edge platforms. This work examines the interplay between edge and cloud deployments, starting from detailed benchmarking of SLM capabilities on single edge devices, and extending to distributed edge clusters. We identify scenarios where edge inference offers comparable performance with lower costs, and others where cloud fallback becomes essential due to limits in scalability or model capacity. Rather than proposing a one-size-fits-all solution, we present platform-level comparisons and design insights for building efficient, adaptive LM inference systems across heterogeneous environments.
Authors: Jiaxin Liu, Jia Wang, Saihui Hou, Min Ren, Huijia Wu, Zhaofeng He
Abstract: In recent years, the rapid development of deepfake technology has given rise to an emerging and serious threat to public security: diffusion-based digital human generation. Unlike traditional face manipulation methods, such models can generate highly realistic videos with consistency through multimodal control signals. Their flexibility and covertness pose severe challenges to existing detection strategies. To bridge this gap, we introduce DigiFakeAV, the new large-scale multimodal digital human forgery dataset based on diffusion models. Employing five of the latest digital human generation methods and the voice cloning methods, we systematically produce a dataset comprising 60,000 videos (8.4 million frames), covering multiple nationalities, skin tones, genders, and real-world scenarios, significantly enhancing data diversity and realism. User studies demonstrate that participants misclassify forged videos as real in 68% of tests, and existing detection models exhibit a large drop in performance on DigiFakeAV, highlighting the challenge of the dataset. To address this problem, we propose DigiShield, an effective detection baseline based on spatiotemporal and cross-modal fusion. By jointly modeling the 3D spatiotemporal features of videos and the semantic-acoustic features of audio, DigiShield achieves state-of-the-art (SOTA) performance on the DigiFakeAV and shows strong generalization on other datasets.
Authors: Ahmed Heakl, Sarim Hashmi, Gustavo Bertolo Stahl, Seung Hun Eddie Han, Salman Khan, Abdulrahman Mahmoud
Abstract: We introduce CASS, the first large-scale dataset and model suite for cross-architecture GPU code transpilation, targeting both source-level (CUDA <--> HIP) and assembly-level (Nvidia SASS <--> AMD RDNA3) translation. The dataset comprises 70k verified code pairs across host and device, addressing a critical gap in low-level GPU code portability. Leveraging this resource, we train the CASS family of domain-specific language models, achieving 95% source translation accuracy and 37.5% assembly translation accuracy, substantially outperforming commercial baselines such as GPT-4o, Claude, and Hipify. Our generated code matches native performance in over 85% of test cases, preserving runtime and memory behavior. To support rigorous evaluation, we introduce CASS-Bench, a curated benchmark spanning 16 GPU domains with ground-truth execution. All data, models, and evaluation tools are released as open source to foster progress in GPU compiler tooling, binary compatibility, and LLM-guided hardware translation.
Authors: Wanghan Xu, Xiangyu Zhao, Yuhao Zhou, Xiaoyu Yue, Ben Fei, Fenghua Ling, Wenlong Zhang, Lei Bai
Abstract: Advancements in Large Language Models (LLMs) drive interest in scientific applications, necessitating specialized benchmarks such as Earth science. Existing benchmarks either present a general science focus devoid of Earth science specificity or cover isolated subdomains, lacking holistic evaluation. Furthermore, current benchmarks typically neglect the assessment of LLMs' capabilities in open-ended scientific exploration. In this paper, we present a comprehensive and professional benchmark for the Earth sciences, designed to evaluate the capabilities of LLMs in scientific exploration within this domain, spanning from fundamental to advanced levels. Leveraging a corpus of 100,000 research papers, we first construct two Question Answering (QA) datasets: Earth-Iron, which offers extensive question coverage for broad assessment, and Earth-Silver, which features a higher level of difficulty to evaluate professional depth. These datasets encompass five Earth spheres, 114 disciplines, and 11 task categories, assessing foundational knowledge crucial for scientific exploration. Most notably, we introduce Earth-Gold with new metrics, a dataset comprising open-ended multi-turn dialogues specifically designed to evaluate the advanced capabilities of LLMs in scientific exploration, including methodology induction, limitation analysis, and concept proposal. Extensive experiments reveal limitations in 11 leading LLMs across different domains and tasks, highlighting considerable room for improvement in their scientific exploration capabilities. The benchmark is available on https://huggingface.co/ai-earth .
Authors: Jiangning Zhu, Yuxing Zhou, Zheng Wang, Juntao Yao, Yima Gu, Yuhui Yuan, Shixia Liu
Abstract: Given the central role of charts in scientific, business, and communication contexts, enhancing the chart understanding capabilities of vision-language models (VLMs) has become increasingly critical. A key limitation of existing VLMs lies in their inaccurate visual grounding of infographic elements, including charts and human-recognizable objects (HROs) such as icons and images. However, chart understanding often requires identifying relevant elements and reasoning over them. To address this limitation, we introduce OrionBench, a benchmark designed to support the development of accurate object detection models for charts and HROs in infographics. It contains 26,250 real and 78,750 synthetic infographics, with over 6.9 million bounding box annotations. These annotations are created by combining the model-in-the-loop and programmatic methods. We demonstrate the usefulness of OrionBench through three applications: 1) constructing a Thinking-with-Boxes scheme to boost the chart understanding performance of VLMs, 2) comparing existing object detection models, and 3) applying the developed detection model to document layout and UI element detection.
Authors: Deyang Kong, Qi Guo, Xiangyu Xi, Wei Wang, Jingang Wang, Xunliang Cai, Shikun Zhang, Wei Ye
Abstract: Reinforcement learning exhibits potential in enhancing the reasoning abilities of large language models, yet it is hard to scale for the low sample efficiency during the rollout phase. Existing methods attempt to improve efficiency by scheduling problems based on problem difficulties. However, these approaches suffer from unstable and biased estimations of problem difficulty and fail to capture the alignment between model competence and problem difficulty in RL training, leading to suboptimal results. To tackle these limitations, this paper introduces $\textbf{C}$ompetence-$\textbf{D}$ifficulty $\textbf{A}$lignment $\textbf{S}$ampling ($\textbf{CDAS}$), which enables accurate and stable estimation of problem difficulties by aggregating historical performance discrepancies of problems. Then the model competence is quantified to adaptively select problems whose difficulty is in alignment with the model's current competence using a fixed-point system. Experimental results across a range of challenging mathematical benchmarks show that CDAS achieves great improvements in both accuracy and efficiency. CDAS attains the highest average accuracy against baselines and exhibits significant speed advantages compared to Dynamic Sampling, a competitive strategy in DAPO, which is 2.33 times slower than CDAS.
Authors: Xueyang Li, Jiahao Li, Yu Song, Yunzhong Lou, Xiangdong Zhou
Abstract: The advent of Computer-Aided Design (CAD) generative modeling will significantly transform the design of industrial products. The recent research endeavor has extended into the realm of Large Language Models (LLMs). In contrast to fine-tuning methods, training-free approaches typically utilize the advanced closed-source LLMs, thereby offering enhanced flexibility and efficiency in the development of AI agents for generating CAD parametric models. However, the substantial cost and limitations of local deployment of the top-tier closed-source LLMs pose challenges in practical applications. The Seek-CAD is the pioneer exploration of locally deployed open-source inference LLM DeepSeek-R1 for CAD parametric model generation with a training-free methodology. This study is the first investigation to incorporate both visual and Chain-of-Thought (CoT) feedback within the self-refinement mechanism for generating CAD models. Specifically, the initial generated parametric CAD model is rendered into a sequence of step-wise perspective images, which are subsequently processed by a Vision Language Model (VLM) alongside the corresponding CoTs derived from DeepSeek-R1 to assess the CAD model generation. Then, the feedback is utilized by DeepSeek-R1 to refine the initial generated model for the next round of generation. Moreover, we present an innovative 3D CAD model dataset structured around the SSR (Sketch, Sketch-based feature, and Refinements) triple design paradigm. This dataset encompasses a wide range of CAD commands, thereby aligning effectively with industrial application requirements and proving suitable for the generation of LLMs. Extensive experiments validate the effectiveness of Seek-CAD under various metrics.
Authors: Pingchuan Ma, Ziang Yin, Qi Jing, Zhengqi Gao, Nicholas Gangi, Boyang Zhang, Tsung-Wei Huang, Zhaoran Huang, Duane S. Boning, Yu Yao, Jiaqi Gu
Abstract: DONNs leverage light propagation for efficient analog AI and signal processing. Advances in nanophotonic fabrication and metasurface-based wavefront engineering have opened new pathways to realize high-capacity DONNs across various spectral regimes. Training such DONN systems to determine the metasurface structures remains challenging. Heuristic methods are fast but oversimplify metasurfaces modulation, often resulting in physically unrealizable designs and significant performance degradation. Simulation-in-the-loop optimizes implementable metasurfaces via adjoint methods, but is computationally prohibitive and unscalable. To address these limitations, we propose SP2RINT, a spatially decoupled, progressive training framework that formulates DONN training as a PDE-constrained learning problem. Metasurface responses are first relaxed into freely trainable transfer matrices with a banded structure. We then progressively enforce physical constraints by alternating between transfer matrix training and adjoint-based inverse design, avoiding per-iteration PDE solves while ensuring final physical realizability. To further reduce runtime, we introduce a physics-inspired, spatially decoupled inverse design strategy based on the natural locality of field interactions. This approach partitions the metasurface into independently solvable patches, enabling scalable and parallel inverse design with system-level calibration. Evaluated across diverse DONN training tasks, SP2RINT achieves digital-comparable accuracy while being 1825 times faster than simulation-in-the-loop approaches. By bridging the gap between abstract DONN models and implementable photonic hardware, SP2RINT enables scalable, high-performance training of physically realizable meta-optical neural systems. Our code is available at https://github.com/ScopeX-ASU/SP2RINT
Authors: Tianyi Ren, Juampablo E. Heras Rivera, Hitender Oswal, Yutong Pan, William Henry, Sophie Walters, Mehmet Kurt
Abstract: Stroke is among the top three causes of death worldwide, and accurate identification of stroke lesion boundaries is critical for diagnosis and treatment. Supervised deep learning methods have emerged as the leading solution for stroke lesion segmentation but require large, diverse, and annotated datasets. The ISLES'24 challenge addresses this need by providing longitudinal stroke imaging data, including CT scans taken on arrival to the hospital and follow-up MRI taken 2-9 days from initial arrival, with annotations derived from follow-up MRI. Importantly, models submitted to the ISLES'24 challenge are evaluated using only CT inputs, requiring prediction of lesion progression that may not be visible in CT scans for segmentation. Our winning solution shows that a carefully designed preprocessing pipeline including deep-learning-based skull stripping and custom intensity windowing is beneficial for accurate segmentation. Combined with a standard large residual nnU-Net architecture for segmentation, this approach achieves a mean test Dice of 28.5 with a standard deviation of 21.27.
Authors: Reva Schwartz, Rumman Chowdhury, Akash Kundu, Heather Frase, Marzieh Fadaee, Tom David, Gabriella Waters, Afaf Taik, Morgan Briggs, Patrick Hall, Shomik Jain, Kyra Yee, Spencer Thomas, Sundeep Bhandari, Qinghua Lu, Matthew Holmes, Theodora Skeadas
Abstract: Conventional AI evaluation approaches concentrated within the AI stack exhibit systemic limitations for exploring, navigating and resolving the human and societal factors that play out in real world deployment such as in education, finance, healthcare, and employment sectors. AI capability evaluations can capture detail about first-order effects, such as whether immediate system outputs are accurate, or contain toxic, biased or stereotypical content, but AI's second-order effects, i.e. any long-term outcomes and consequences that may result from AI use in the real world, have become a significant area of interest as the technology becomes embedded in our daily lives. These secondary effects can include shifts in user behavior, societal, cultural and economic ramifications, workforce transformations, and long-term downstream impacts that may result from a broad and growing set of risks. This position paper argues that measuring the indirect and secondary effects of AI will require expansion beyond static, single-turn approaches conducted in silico to include testing paradigms that can capture what actually materializes when people use AI technology in context. Specifically, we describe the need for data and methods that can facilitate contextual awareness and enable downstream interpretation and decision making about AI's secondary effects, and recommend requirements for a new ecosystem.
Authors: Ashirbad Mishra, Jinyu Zhao, Soumik Dey, Hansi Wu, Binbin Li, Kamesh Madduri
Abstract: In the domain of sponsored search advertising, the focus of Keyphrase recommendation has largely been on exact match types, which pose issues such as high management expenses, limited targeting scope, and evolving search query patterns. Alternatives like Broad match types can alleviate certain drawbacks of exact matches but present challenges like poor targeting accuracy and minimal supervisory signals owing to limited advertiser usage. This research defines the criteria for an ideal broad match, emphasizing on both efficiency and effectiveness, ensuring that a significant portion of matched queries are relevant. We propose BroadGen, an innovative framework that recommends efficient and effective broad match keyphrases by utilizing historical search query data. Additionally, we demonstrate that BroadGen, through token correspondence modeling, maintains better query stability over time. BroadGen's capabilities allow it to serve daily, millions of sellers at eBay with over 2.3 billion items.
Authors: Ken Huang, Vineeth Sai Narajala, John Yeoh, Jason Ross, Ramesh Raskar, Youssef Harkati, Jerry Huang, Idan Habler, Chris Hughes
Abstract: Traditional Identity and Access Management (IAM) systems, primarily designed for human users or static machine identities via protocols such as OAuth, OpenID Connect (OIDC), and SAML, prove fundamentally inadequate for the dynamic, interdependent, and often ephemeral nature of AI agents operating at scale within Multi Agent Systems (MAS), a computational system composed of multiple interacting intelligent agents that work collectively. This paper posits the imperative for a novel Agentic AI IAM framework: We deconstruct the limitations of existing protocols when applied to MAS, illustrating with concrete examples why their coarse-grained controls, single-entity focus, and lack of context-awareness falter. We then propose a comprehensive framework built upon rich, verifiable Agent Identities (IDs), leveraging Decentralized Identifiers (DIDs) and Verifiable Credentials (VCs), that encapsulate an agents capabilities, provenance, behavioral scope, and security posture. Our framework includes an Agent Naming Service (ANS) for secure and capability-aware discovery, dynamic fine-grained access control mechanisms, and critically, a unified global session management and policy enforcement layer for real-time control and consistent revocation across heterogeneous agent communication protocols. We also explore how Zero-Knowledge Proofs (ZKPs) enable privacy-preserving attribute disclosure and verifiable policy compliance. We outline the architecture, operational lifecycle, innovative contributions, and security considerations of this new IAM paradigm, aiming to establish the foundational trust, accountability, and security necessary for the burgeoning field of agentic AI and the complex ecosystems they will inhabit.
Authors: Bingdong Li, Mei Jiang, Hong Qian, Ke Tang, Aimin Zhou, Peng Yang
Abstract: Evolutionary Reinforcement Learning (ERL), training the Reinforcement Learning (RL) policies with Evolutionary Algorithms (EAs), have demonstrated enhanced exploration capabilities and greater robustness than using traditional policy gradient. However, ERL suffers from the high computational costs and low search efficiency, as EAs require evaluating numerous candidate policies with expensive simulations, many of which are ineffective and do not contribute meaningfully to the training. One intuitive way to reduce the ineffective evaluations is to adopt the surrogates. Unfortunately, existing ERL policies are often modeled as deep neural networks (DNNs) and thus naturally represented as high-dimensional vectors containing millions of weights, which makes the building of effective surrogates for ERL policies extremely challenging. This paper proposes a novel surrogate-assisted ERL that integrates Autoencoders (AE) and Hyperbolic Neural Networks (HNN). Specifically, AE compresses high-dimensional policies into low-dimensional representations while extracting key features as the inputs for the surrogate. HNN, functioning as a classification-based surrogate model, can learn complex nonlinear relationships from sampled data and enable more accurate pre-selection of the sampled policies without real evaluations. The experiments on 10 Atari and 4 Mujoco games have verified that the proposed method outperforms previous approaches significantly. The search trajectories guided by AE and HNN are also visually demonstrated to be more effective, in terms of both exploration and convergence. This paper not only presents the first learnable policy embedding and surrogate-modeling modules for high-dimensional ERL policies, but also empirically reveals when and why they can be successful.
Authors: Lijun Zhang, Lin Li, Yajie Qi, Huizhong Song, Yaodong Yang, Jun Wang, Wei Wei
Abstract: When fine-tuning pre-trained Large Language Models (LLMs) to align with human values and intentions, maximizing the estimated reward can lead to superior performance, but it also introduces potential risks due to deviations from the reference model's intended behavior. Most existing methods typically introduce KL divergence to constrain deviations between the trained model and the reference model; however, this may not be sufficient in certain applications that require tight risk control. In this paper, we introduce Risk-aware Direct Preference Optimization (Ra-DPO), a novel approach that incorporates risk-awareness by employing a class of nested risk measures. This approach formulates a constrained risk-aware advantage function maximization problem and then converts the Bradley-Terry model into a token-level representation. The objective function maximizes the likelihood of the policy while suppressing the deviation between a trained model and the reference model using a sequential risk ratio, thereby enhancing the model's risk-awareness. Experimental results across three open-source datasets: IMDb Dataset, Anthropic HH Dataset, and AlpacaEval, demonstrate the proposed method's superior performance in balancing alignment performance and model drift. Our code is opensourced at https://github.com/zlj123-max/Ra-DPO.
Authors: Jihoon Lee, Min Song
Abstract: Despite significant advancements in Large Vision-Language Models, Object Hallucination (OH) remains a persistent challenge. Building upon prior studies on contrastive decoding that address this issue without requiring additional model training, we introduce RVCD (Retrieval Visual Contrastive Decoding), an advanced method to suppress OH. RVCD leverages both negative and positive images at the logit level, explicitly referencing AI-generated images designed to represent a single concept. Our approach demonstrates substantial improvements over existing decoding-based methods.
Authors: Dane Malenfant, Blake A. Richards
Abstract: Sometimes we benefit from actions that others have taken even when we are unaware that they took those actions. For example, if your neighbor chooses not to take a parking spot in front of your house when you are not there, you can benefit, even without being aware that they took this action. These "hidden gifts" represent an interesting challenge for multi-agent reinforcement learning (MARL), since assigning credit when the beneficial actions of others are hidden is non-trivial. Here, we study the impact of hidden gifts with a very simple MARL task. In this task, agents in a grid-world environment have individual doors to unlock in order to obtain individual rewards. As well, if all the agents unlock their door the group receives a larger collective reward. However, there is only one key for all of the doors, such that the collective reward can only be obtained when the agents drop the key for others after they use it. Notably, there is nothing to indicate to an agent that the other agents have dropped the key, thus the act of dropping the key for others is a "hidden gift". We show that several different state-of-the-art RL algorithms, including MARL algorithms, fail to learn how to obtain the collective reward in this simple task. Interestingly, we find that independent model-free policy gradient agents can solve the task when we provide them with information about their own action history, but MARL agents still cannot solve the task with action history. Finally, we derive a correction term for these independent agents, inspired by learning aware approaches, which reduces the variance in learning and helps them to converge to collective success more reliably. These results show that credit assignment in multi-agent settings can be particularly challenging in the presence of "hidden gifts", and demonstrate that learning awareness in independent agents can benefit these settings.
Authors: Injae Na, Keonwoong Noh, Woohwan Jung
Abstract: LLM providers typically offer multiple LLM tiers, varying in performance and price. As NLP tasks become more complex and modularized, selecting the suitable LLM tier for each subtask is a key challenge to balance between cost and performance. To address the problem, we introduce LLM Automatic Transmission (LLM-AT) framework that automatically selects LLM tiers without training. LLM-AT consists of Starter, Generator, and Judge. The starter selects the initial LLM tier expected to solve the given question, the generator produces a response using the LLM of the selected tier, and the judge evaluates the validity of the response. If the response is invalid, LLM-AT iteratively upgrades to a higher-tier model, generates a new response, and re-evaluates until a valid response is obtained. Additionally, we propose accuracy estimator, which enables the suitable initial LLM tier selection without training. Given an input question, accuracy estimator estimates the expected accuracy of each LLM tier by computing the valid response rate across top-k similar queries from past inference records. Experiments demonstrate that LLM-AT achieves superior performance while reducing costs, making it a practical solution for real-world applications.
Authors: Shiqi Zhang, Tuomas Virtanen
Abstract: Bioacoustic sound event detection (BioSED) is crucial for biodiversity conservation but faces practical challenges during model development and training: limited amounts of annotated data, sparse events, species diversity, and class imbalance. To address these challenges efficiently with a limited labeling budget, we apply the mismatch-first farthest-traversal (MFFT), an active learning method integrating committee voting disagreement and diversity analysis. We also refine an existing BioSED dataset specifically for evaluating active learning algorithms. Experimental results demonstrate that MFFT achieves a mAP of 68% when cold-starting and 71% when warm-starting (which is close to the fully-supervised mAP of 75%) while using only 2.3% of the annotations. Notably, MFFT excels in cold-start scenarios and with rare species, which are critical for monitoring endangered species, demonstrating its practical value.
Authors: Anissa Alloula, Charles Jones, Ben Glocker, Bart{\l}omiej W. Papie\.z
Abstract: Despite the constant development of new bias mitigation methods for machine learning, no method consistently succeeds, and a fundamental question remains unanswered: when and why do bias mitigation techniques fail? In this paper, we hypothesise that a key factor may be the often-overlooked but crucial step shared by many bias mitigation methods: the definition of subgroups. To investigate this, we conduct a comprehensive evaluation of state-of-the-art bias mitigation methods across multiple vision and language classification tasks, systematically varying subgroup definitions, including coarse, fine-grained, intersectional, and noisy subgroups. Our results reveal that subgroup choice significantly impacts performance, with certain groupings paradoxically leading to worse outcomes than no mitigation at all. Our findings suggest that observing a disparity between a set of subgroups is not a sufficient reason to use those subgroups for mitigation. Through theoretical analysis, we explain these phenomena and uncover a counter-intuitive insight that, in some cases, improving fairness with respect to a particular set of subgroups is best achieved by using a different set of subgroups for mitigation. Our work highlights the importance of careful subgroup definition in bias mitigation and presents it as an alternative lever for improving the robustness and fairness of machine learning models.
Authors: Jingyuan Huang, Xi Zhu, Minghao Guo, Yongfeng Zhang
Abstract: Web 2.0 social platforms are inherently centralized, with user data and algorithmic decisions controlled by the platform. However, users can only passively receive social predictions without being able to choose the underlying algorithm, which limits personalization. Fortunately, with the emergence of blockchain, users are allowed to choose algorithms that are tailored to their local situation, improving prediction results in a personalized way. In a blockchain environment, each user possesses its own model to perform the social prediction, capturing different perspectives on social interactions. In our work, we propose DeSocial, a decentralized social network learning framework deployed on an Ethereum (ETH) local development chain that integrates distributed data storage, node-level consensus, and user-driven model selection through Ganache. In the first stage, each user leverages DeSocial to evaluate multiple backbone models on their local subgraph. DeSocial coordinates the execution and returns model-wise prediction results, enabling the user to select the most suitable backbone for personalized social prediction. Then, DeSocial uniformly selects several validation nodes that possess the algorithm specified by each user, and aggregates the prediction results by majority voting, to prevent errors caused by any single model's misjudgment. Extensive experiments show that DeSocial has an evident improvement compared to the five classical centralized social network learning models, promoting user empowerment in blockchain-based decentralized social networks, showing the importance of multi-node validation and personalized algorithm selection based on blockchain. Our implementation is available at: https://github.com/agiresearch/DeSocial.
Authors: Haoming Song, Delin Qu, Yuanqi Yao, Qizhi Chen, Qi Lv, Yiwen Tang, Modi Shi, Guanghui Ren, Maoqing Yao, Bin Zhao, Dong Wang, Xuelong Li
Abstract: Humans practice slow thinking before performing actual actions when handling complex tasks in the physical world. This thinking paradigm, recently, has achieved remarkable advancement in boosting Large Language Models (LLMs) to solve complex tasks in digital domains. However, the potential of slow thinking remains largely unexplored for robotic foundation models interacting with the physical world. In this work, we propose Hume: a dual-system Vision-Language-Action (VLA) model with value-guided System-2 thinking and cascaded action denoising, exploring human-like thinking capabilities of Vision-Language-Action models for dexterous robot control. System 2 of Hume implements value-Guided thinking by extending a Vision-Language-Action Model backbone with a novel value-query head to estimate the state-action value of predicted actions. The value-guided thinking is conducted by repeat sampling multiple action candidates and selecting one according to state-action value. System 1 of Hume is a lightweight reactive visuomotor policy that takes System 2 selected action and performs cascaded action denoising for dexterous robot control. At deployment time, System 2 performs value-guided thinking at a low frequency while System 1 asynchronously receives the System 2 selected action candidate and predicts fluid actions in real time. We show that Hume outperforms the existing state-of-the-art Vision-Language-Action models across multiple simulation benchmark and real-robot deployments.
Authors: Jianheng Zhuo, Yifan Yang, Yiwen Shao, Yong Xu, Dong Yu, Kai Yu, Xie Chen
Abstract: Automatic speech recognition (ASR) has made remarkable progress but heavily relies on large-scale labeled data, which is scarce for low-resource languages like Vietnamese. While existing systems such as Whisper, USM, and MMS achieve promising performance, their efficacy remains inadequate in terms of training costs, latency, and accessibility. To address these issues, we propose VietASR, a novel ASR training pipeline that leverages vast amounts of unlabeled data and a small set of labeled data. Through multi-iteration ASR-biased self-supervised learning on a large-scale unlabeled dataset, VietASR offers a cost-effective and practical solution for enhancing ASR performance. Experiments demonstrate that pre-training on 70,000-hour unlabeled data and fine-tuning on merely 50-hour labeled data yield a lightweight but powerful ASR model. It outperforms Whisper Large-v3 and commercial ASR systems on real-world data. Our code and models will be open-sourced to facilitate research in low-resource ASR.
Authors: Yuxin Ren, Maxwell D Collins, Miao Hu, Huanrui Yang
Abstract: While transformers excel across vision and language pretraining tasks, their reliance on attention mechanisms poses challenges for inference efficiency, especially on edge and embedded accelerators with limited parallelism and memory bandwidth. Hinted by the observed redundancy of attention at inference time, we hypothesize that though the model learns complicated token dependency through pretraining, the inference-time sequence-to-sequence mapping in each attention layer is actually ''simple'' enough to be represented with a much cheaper function. In this work, we explore FAR, a Function-preserving Attention Replacement framework that replaces all attention blocks in pretrained transformers with learnable sequence-to-sequence modules, exemplified by an LSTM. FAR optimize a multi-head LSTM architecture with a block-wise distillation objective and a global structural pruning framework to achieve a family of efficient LSTM-based models from pretrained transformers. We validate FAR on the DeiT vision transformer family and demonstrate that it matches the accuracy of the original models on ImageNet and multiple downstream tasks with reduced parameters and latency. Further analysis shows that FAR preserves the semantic token relationships and the token-to-token correlation learned in the transformer's attention module.
Authors: M\'onika Farsang, Ramin Hasani, Radu Grosu
Abstract: We present LrcSSM, a $\textit{nonlinear}$ recurrent model that processes long sequences as fast as today's linear state-space layers. By forcing the state-transition matrix to be diagonal and learned at every step, the full sequence can be solved in parallel with a single prefix-scan, giving $\mathcal{O}(TD)$ time and memory and only $\mathcal{O}(\log T)$ sequential depth, for input-sequence length $T$ and a state dimension $D$. Moreover, LrcSSM offers a formal gradient-stability guarantee that other input-varying systems such as Liquid-S4 and Mamba do not provide. Lastly, for network depth $L$, as the forward and backward passes cost $\Theta(T\,D\,L)$ FLOPs, with its low sequential depth and parameter count $\Theta(D\,L)$, the model follows the compute-optimal scaling law regime ($\beta \approx 0.42$) recently observed for Mamba, outperforming quadratic-attention Transformers at equal compute while avoiding the memory overhead of FFT-based long convolutions. We show that on a series of long-range forecasting tasks, LrcSSM outperforms LRU, S5 and Mamba.
Authors: Pardis Taghavi, Tian Liu, Renjie Li, Reza Langari, Zhengzhong Tu
Abstract: Instance segmentation demands costly per-pixel annotations and large models. We introduce CAST, a semi-supervised knowledge distillation (SSKD) framework that compresses pretrained vision foundation models (VFM) into compact experts using limited labeled and abundant unlabeled data. CAST unfolds in three stages: (1) domain adaptation of the VFM teacher(s) via self-training with contrastive pixel calibration, (2) distillation into a compact student via a unified multi-objective loss that couples standard supervision and pseudo-labels with our instance-aware pixel-wise contrastive term, and (3) fine-tuning on labeled data to remove residual pseudo-label bias. Central to CAST is an \emph{instance-aware pixel-wise contrastive loss} that fuses mask and class scores to mine informative negatives and enforce clear inter-instance margins. By maintaining this contrastive signal across both adaptation and distillation, we align teacher and student embeddings and fully leverage unlabeled images. On Cityscapes and ADE20K, our ~11X smaller student surpasses its adapted VFM teacher(s) by +3.4 AP (33.9 vs. 30.5) and +1.5 AP (16.7 vs. 15.2) and outperforms state-of-the-art semi-supervised approaches.
Authors: Mengdan Zhu, Senhao Cheng, Guangji Bai, Yifei Zhang, Liang Zhao
Abstract: Text-to-image generation increasingly demands access to domain-specific, fine-grained, and rapidly evolving knowledge that pretrained models cannot fully capture. Existing Retrieval-Augmented Generation (RAG) methods attempt to address this by retrieving globally relevant images, but they fail when no single image contains all desired elements from a complex user query. We propose Cross-modal RAG, a novel framework that decomposes both queries and images into sub-dimensional components, enabling subquery-aware retrieval and generation. Our method introduces a hybrid retrieval strategy - combining a sub-dimensional sparse retriever with a dense retriever - to identify a Pareto-optimal set of images, each contributing complementary aspects of the query. During generation, a multimodal large language model is guided to selectively condition on relevant visual features aligned to specific subqueries, ensuring subquery-aware image synthesis. Extensive experiments on MS-COCO, Flickr30K, WikiArt, CUB, and ImageNet-LT demonstrate that Cross-modal RAG significantly outperforms existing baselines in both retrieval and generation quality, while maintaining high efficiency.
Authors: Tianjun Gu, Linfeng Li, Xuhong Wang, Chenghua Gong, Jingyu Gong, Zhizhong Zhang, Yuan Xie, Lizhuang Ma, Xin Tan
Abstract: Adaptive navigation in unfamiliar environments is crucial for household service robots but remains challenging due to the need for both low-level path planning and high-level scene understanding. While recent vision-language model (VLM) based zero-shot approaches reduce dependence on prior maps and scene-specific training data, they face significant limitations: spatiotemporal discontinuity from discrete observations, unstructured memory representations, and insufficient task understanding leading to navigation failures. We propose DORAEMON (Decentralized Ontology-aware Reliable Agent with Enhanced Memory Oriented Navigation), a novel cognitive-inspired framework consisting of Ventral and Dorsal Streams that mimics human navigation capabilities. The Dorsal Stream implements the Hierarchical Semantic-Spatial Fusion and Topology Map to handle spatiotemporal discontinuities, while the Ventral Stream combines RAG-VLM and Policy-VLM to improve decision-making. Our approach also develops Nav-Ensurance to ensure navigation safety and efficiency. We evaluate DORAEMON on the HM3D, MP3D, and GOAT datasets, where it achieves state-of-the-art performance on both success rate (SR) and success weighted by path length (SPL) metrics, significantly outperforming existing methods. We also introduce a new evaluation metric (AORI) to assess navigation intelligence better. Comprehensive experiments demonstrate DORAEMON's effectiveness in zero-shot autonomous navigation without requiring prior map building or pre-training.
Authors: Yudi Zhang, Weilin Zhao, Xu Han, Tiejun Zhao, Wang Xu, Hailong Cao, Conghui Zhu
Abstract: Speculative decoding and quantization effectively accelerate memory-bound inference of large language models. Speculative decoding mitigates the memory bandwidth bottleneck by verifying multiple tokens within a single forward pass, which increases computational effort. Quantization achieves this optimization by compressing weights and activations into lower bit-widths and also reduces computations via low-bit matrix multiplications. To further leverage their strengths, we investigate the integration of these two techniques. Surprisingly, experiments applying the advanced speculative decoding method EAGLE-2 to various quantized models reveal that the memory benefits from 4-bit weight quantization are diminished by the computational load from speculative decoding. Specifically, verifying a tree-style draft incurs significantly more time overhead than a single-token forward pass on 4-bit weight quantized models. This finding led to our new speculative decoding design: a hierarchical framework that employs a small model as an intermediate stage to turn tree-style drafts into sequence drafts, leveraging the memory access benefits of the target quantized model. Experimental results show that our hierarchical approach achieves a 2.78$\times$ speedup across various tasks for the 4-bit weight Llama-3-70B model on an A100 GPU, outperforming EAGLE-2 by 1.31$\times$. Code available at https://github.com/AI9Stars/SpecMQuant.
Authors: Jujie He, Jiacai Liu, Chris Yuhao Liu, Rui Yan, Chaojie Wang, Peng Cheng, Xiaoyu Zhang, Fuxiang Zhang, Jiacheng Xu, Wei Shen, Siyuan Li, Liang Zeng, Tianwen Wei, Cheng Cheng, Bo An, Yang Liu, Yahui Zhou
Abstract: The success of DeepSeek-R1 underscores the significant role of reinforcement learning (RL) in enhancing the reasoning capabilities of large language models (LLMs). In this work, we present Skywork-OR1, an effective and scalable RL implementation for long Chain-of-Thought (CoT) models. Building on the DeepSeek-R1-Distill model series, our RL approach achieves notable performance gains, increasing average accuracy across AIME24, AIME25, and LiveCodeBench from 57.8% to 72.8% (+15.0%) for the 32B model and from 43.6% to 57.5% (+13.9%) for the 7B model. Our Skywork-OR1-32B model surpasses both DeepSeek-R1 and Qwen3-32B on the AIME24 and AIME25 benchmarks, while achieving comparable results on LiveCodeBench. The Skywork-OR1-7B and Skywork-OR1-Math-7B models demonstrate competitive reasoning capabilities among models of similar size. We perform comprehensive ablation studies on the core components of our training pipeline to validate their effectiveness. Additionally, we thoroughly investigate the phenomenon of entropy collapse, identify key factors affecting entropy dynamics, and demonstrate that mitigating premature entropy collapse is critical for improved test performance. To support community research, we fully open-source our model weights, training code, and training datasets.
Authors: Haomiao Qiu, Miao Zhang, Ziyue Qiao, Weili Guan, Min Zhang, Liqiang Nie
Abstract: Continual Learning requires a model to learn multiple tasks in sequence while maintaining both stability:preserving knowledge from previously learned tasks, and plasticity:effectively learning new tasks. Gradient projection has emerged as an effective and popular paradigm in CL, where it partitions the gradient space of previously learned tasks into two orthogonal subspaces: a primary subspace and a minor subspace. New tasks are learned effectively within the minor subspace, thereby reducing interference with previously acquired knowledge. However, existing Gradient Projection methods struggle to achieve an optimal balance between plasticity and stability, as it is hard to appropriately partition the gradient space. In this work, we consider a continual learning paradigm based on Low-Rank Adaptation, which has gained considerable attention due to its efficiency and wide applicability, and propose a novel approach for continual learning, called SplitLoRA. We first provide a theoretical analysis of how subspace partitioning affects model stability and plasticity. Informed by this analysis, we then introduce an effective method that derives the optimal partition of the gradient space for previously learned tasks. This approach effectively balances stability and plasticity in continual learning. Experimental results on multiple datasets demonstrate that the proposed method achieves state-of-the-art performance.
Authors: Haomiao Qiu, Miao Zhang, Ziyue Qiao, Liqiang Nie
Abstract: Continual Learning (CL) aims to enable models to continuously acquire new knowledge from a sequence of tasks with avoiding the forgetting of learned information. However, existing CL methods only rely on the parameters of the most recent task for inference, which makes them susceptible to catastrophic forgetting. Inspired by the recent success of model merging techniques, we propose \textbf{Perturb-and-Merge (P\&M)}, a novel continual learning framework that integrates model merging into the CL paradigm to mitigate forgetting. Specifically, after training on each task, P\&M constructs a new model by forming a convex combination of the previous model and the newly trained task-specific model. Through theoretical analysis, we minimize the total loss increase across all tasks and derive an analytical solution for the optimal merging coefficient. To further improve the performance of the merged model, we observe that the degradation introduced during merging can be alleviated by a regularization term composed of the task vector and the Hessian matrix of the loss function. Interestingly, we show that this term can be efficiently approximated using second-order symmetric finite differences, and a stochastic perturbation strategy along the task vector direction is accordingly devised which incurs no additional forward or backward passes while providing an effective approximation of the regularization term. Finally, we combine P\&M with LoRA, a parameter-efficient fine-tuning method, to reduce memory overhead. Our proposed approach achieves state-of-the-art performance on several continual learning benchmark datasets.
Authors: Jiaxi Yang, Mengqi Zhang, Yiqiao Jin, Hao Chen, Qingsong Wen, Lu Lin, Yi He, Weijie Xu, James Evans, Jindong Wang
Abstract: Large Language Model-based Multi-Agent Systems (MASs) have emerged as a powerful paradigm for tackling complex tasks through collaborative intelligence. Nevertheless, the question of how agents should be structurally organized for optimal cooperation remains largely unexplored. In this position paper, we aim to gently redirect the focus of the MAS research community toward this critical dimension: develop topology-aware MASs for specific tasks. Specifically, the system consists of three core components - agents, communication links, and communication patterns - that collectively shape its coordination performance and efficiency. To this end, we introduce a systematic, three-stage framework: agent selection, structure profiling, and topology synthesis. Each stage would trigger new research opportunities in areas such as language models, reinforcement learning, graph learning, and generative modeling; together, they could unleash the full potential of MASs in complicated real-world applications. Then, we discuss the potential challenges and opportunities in the evaluation of multiple systems. We hope our perspective and framework can offer critical new insights in the era of agentic AI.
Authors: Hoang Pham, Thuy-Duong Nguyen, Khac-Hoai Nam Bui
Abstract: This paper presents a novel approach for unified retrieval-augmented generation (RAG) systems using the recent emerging large language model (LLM) agent concept. Specifically, Agent LLM, which utilizes LLM as fundamental controllers, has become a promising approach to enable the interpretability of RAG tasks, especially for complex reasoning question-answering systems (e.g., multi-hop queries). Nonetheless, previous works mainly focus on solving RAG systems with either single-hop or multi-hop approaches separately, which limits the application of those approaches to real-world applications. In this study, we propose a trainable agent framework called Agent-UniRAG for unified retrieval-augmented LLM systems, which enhances the effectiveness and interpretability of RAG systems. The main idea is to design an LLM agent framework to solve RAG tasks step-by-step based on the complexity of the inputs, simultaneously including single-hop and multi-hop queries in an end-to-end manner. Furthermore, we introduce SynAgent-RAG, a synthetic dataset to enable the proposed agent framework for small open-source LLMs (e.g., Llama-3-8B). The results show comparable performances with closed-source and larger open-source LLMs across various RAG benchmarks. Our source code and dataset are publicly available for further exploitation.
Authors: Luca Maria Del Bono, Federico Ricci-Tersenghi, Francesco Zamponi
Abstract: Recent years have seen a rise in the application of machine learning techniques to aid the simulation of hard-to-sample systems that cannot be studied using traditional methods. Despite the introduction of many different architectures and procedures, a wide theoretical understanding is still lacking, with the risk of suboptimal implementations. As a first step to address this gap, we provide here a complete analytic study of the widely-used Sequential Tempering procedure applied to a shallow MADE architecture for the Curie-Weiss model. The contribution of this work is twofold: firstly, we give a description of the optimal weights and of the training under Gradient Descent optimization. Secondly, we compare what happens in Sequential Tempering with and without the addition of local Metropolis Monte Carlo steps. We are thus able to give theoretical predictions on the best procedure to apply in this case. This work establishes a clear theoretical basis for the integration of machine learning techniques into Monte Carlo sampling and optimization.
Authors: Younggyo Seo, Carmelo Sferrazza, Haoran Geng, Michal Nauman, Zhao-Heng Yin, Pieter Abbeel
Abstract: Reinforcement learning (RL) has driven significant progress in robotics, but its complexity and long training times remain major bottlenecks. In this report, we introduce FastTD3, a simple, fast, and capable RL algorithm that significantly speeds up training for humanoid robots in popular suites such as HumanoidBench, IsaacLab, and MuJoCo Playground. Our recipe is remarkably simple: we train an off-policy TD3 agent with several modifications -- parallel simulation, large-batch updates, a distributional critic, and carefully tuned hyperparameters. FastTD3 solves a range of HumanoidBench tasks in under 3 hours on a single A100 GPU, while remaining stable during training. We also provide a lightweight and easy-to-use implementation of FastTD3 to accelerate RL research in robotics.
Authors: Guoxuan Chen, Lianghao Xia, Chao Huang
Abstract: Modern recommender systems powered by Graph Neural Networks (GNNs) excel at modeling complex user-item interactions, yet increasingly face scenarios requiring selective forgetting of training data. Beyond user requests to remove specific interactions due to privacy concerns or preference changes, regulatory frameworks mandate recommender systems' ability to eliminate the influence of certain user data from models. This recommendation unlearning challenge presents unique difficulties as removing connections within interaction graphs creates ripple effects throughout the model, potentially impacting recommendations for numerous users. Traditional approaches suffer from significant drawbacks: fragmentation methods damage graph structure and diminish performance, while influence function techniques make assumptions that may not hold in complex GNNs, particularly with self-supervised or random architectures. To address these limitations, we propose a novel model-agnostic pre-training paradigm UnlearnRec that prepares systems for efficient unlearning operations. Our Influence Encoder takes unlearning requests together with existing model parameters and directly produces updated parameters of unlearned model with little fine-tuning, avoiding complete retraining while preserving model performance characteristics. Extensive evaluation on public benchmarks demonstrates that our method delivers exceptional unlearning effectiveness while providing more than 10x speedup compared to retraining approaches. We release our method implementation at: https://github.com/HKUDS/UnlearnRec.
Authors: Mihir Prabhudesai, Lili Chen, Alex Ippoliti, Katerina Fragkiadaki, Hao Liu, Deepak Pathak
Abstract: Reinforcement learning (RL) has enabled machine learning models to achieve significant advances in many fields. Most recently, RL has empowered frontier language models to solve challenging math, science, and coding problems. However, central to any RL algorithm is the reward function, and reward engineering is a notoriously difficult problem in any domain. In this paper, we propose RENT: Reinforcement Learning via Entropy Minimization -- a fully unsupervised RL method that requires no external reward or ground-truth answers, and instead uses the model's entropy of its underlying distribution as an intrinsic reward. We find that by reinforcing the chains of thought that yield high model confidence on its generated answers, the model improves its reasoning ability. In our experiments, we showcase these improvements on an extensive suite of commonly-used reasoning benchmarks, including GSM8K, MATH500, AMC, AIME, and GPQA, and models of varying sizes from the Qwen and Mistral families. The generality of our unsupervised learning method lends itself to applicability in a wide range of domains where external supervision is unavailable.