Authors: Caroline Wang, Daniel Kasenberg, Kim Stachenfeld, Pablo Samuel Castro
Abstract: As Large Language Models (LLMs) are increasingly deployed in social and strategic scenarios, it becomes critical to understand where and why their behavior diverges from that of humans. While behavioral game theory (BGT) provides a framework for analyzing behavior, existing models do not fully capture the idiosyncratic behavior of humans or black-box, non-human agents like LLMs. We employ AlphaEvolve, a cutting-edge program discovery tool, to directly discover interpretable models of human and LLM behavior from data, thereby enabling open-ended discovery of structural factors driving human and LLM behavior. Our analysis on iterated rock-paper-scissors reveals that frontier LLMs can be capable of deeper strategic behavior than humans. These results provide a foundation for understanding structural differences driving differences in human and LLM behavior in strategic interactions.
Authors: Zhiling Yan, Dingjie Song, Zhe Fang, Yisheng Ji, Xiang Li, Quanzheng Li, Lichao Sun
Abstract: The deployment of Large Language Models (LLMs) in high-stakes clinical settings demands rigorous and reliable evaluation. However, existing medical benchmarks remain static, suffering from two critical limitations: (1) data contamination, where test sets inadvertently leak into training corpora, leading to inflated performance estimates; and (2) temporal misalignment, failing to capture the rapid evolution of medical knowledge. Furthermore, current evaluation metrics for open-ended clinical reasoning often rely on either shallow lexical overlap (e.g., ROUGE) or subjective LLM-as-a-Judge scoring, both inadequate for verifying clinical correctness. To bridge these gaps, we introduce LiveMedBench, a continuously updated, contamination-free, and rubric-based benchmark that weekly harvests real-world clinical cases from online medical communities, ensuring strict temporal separation from model training data. We propose a Multi-Agent Clinical Curation Framework that filters raw data noise and validates clinical integrity against evidence-based medical principles. For evaluation, we develop an Automated Rubric-based Evaluation Framework that decomposes physician responses into granular, case-specific criteria, achieving substantially stronger alignment with expert physicians than LLM-as-a-Judge. To date, LiveMedBench comprises 2,756 real-world cases spanning 38 medical specialties and multiple languages, paired with 16,702 unique evaluation criteria. Extensive evaluation of 38 LLMs reveals that even the best-performing model achieves only 39.2%, and 84% of models exhibit performance degradation on post-cutoff cases, confirming pervasive data contamination risks. Error analysis further identifies contextual application-not factual knowledge-as the dominant bottleneck, with 35-48% of failures stemming from the inability to tailor medical knowledge to patient-specific constraints.
Authors: Yansong Qu, Zihao Sheng, Zilin Huang, Jiancong Chen, Yuhao Luo, Tianyi Wang, Yiheng Feng, Samuel Labi, Sikai Chen
Abstract: Reinforcement Learning (RL) has emerged as a dominant paradigm for end-to-end autonomous driving (AD). However, RL suffers from sample inefficiency and a lack of semantic interpretability in complex scenarios. Foundation Models, particularly Vision-Language Models (VLMs), can mitigate this by offering rich, context-aware knowledge, yet their high inference latency hinders deployment in high-frequency RL training loops. To bridge this gap, we present Found-RL, a platform tailored to efficiently enhance RL for AD using foundation models. A core innovation is the asynchronous batch inference framework, which decouples heavy VLM reasoning from the simulation loop, effectively resolving latency bottlenecks to support real-time learning. We introduce diverse supervision mechanisms: Value-Margin Regularization (VMR) and Advantage-Weighted Action Guidance (AWAG) to effectively distill expert-like VLM action suggestions into the RL policy. Additionally, we adopt high-throughput CLIP for dense reward shaping. We address CLIP's dynamic blindness via Conditional Contrastive Action Alignment, which conditions prompts on discretized speed/command and yields a normalized, margin-based bonus from context-specific action-anchor scoring. Found-RL provides an end-to-end pipeline for fine-tuned VLM integration and shows that a lightweight RL model can achieve near-VLM performance compared with billion-parameter VLMs while sustaining real-time inference (approx. 500 FPS). Code, data, and models will be publicly available at https://github.com/ys-qu/found-rl.
Authors: Jihwan Oh, Murad Aghazada, Yooju Shin, Se-Young Yun, Taehyeon Kim
Abstract: Bargaining is often regarded as a logical arena rather than an art or a matter of intuition, yet Large Language Models (LLMs) still struggle to navigate it due to limited strategic depth and difficulty adapting to complex human factors. Current benchmarks rarely capture this limitation. To bridge this gap, we present an utility feedback centric framework. Our contributions are: (i) AgoraBench, a new benchmark spanning nine challenging settings (e.g., deception, monopoly) that supports diverse strategy modeling; (ii) human-aligned, economically grounded metrics derived from utility theory. This is operationalized via agent utility, negotiation power, and acquisition ratio that implicitly measure how well the negotiation aligns with human preference and (iii) a human preference grounded dataset with learning pipeline that strengthens LLMs' bargaining ability through both prompting and finetuning. Empirical results indicate that baseline LLM strategies often diverge from human preferences, while our mechanism substantially improves negotiation performance, yielding deeper strategic behavior and stronger opponent awareness.
Authors: Zhenhe Cui, Huaxiang Xia, Hangjun Shen, Kailun Luo, Yong He, Wei Liang
Abstract: Qualitative Numerical Planning (QNP) serves as an important abstraction model for generalized planning (GP), which aims to compute general plans that solve multiple instances at once. Recent works show that large language models (LLMs) can function as generalized planners. This work investigates whether LLMs can serve as QNP abstraction generators for GP problems and how to fix abstractions via automated debugging. We propose a prompt protocol: input a GP domain and training tasks to LLMs, prompting them to generate abstract features and further abstract the initial state, action set, and goal into QNP problems. An automated debugging method is designed to detect abstraction errors, guiding LLMs to fix abstractions. Experiments demonstrate that under properly guided by automated debugging, some LLMs can generate useful QNP abstractions.
Authors: Bo Xue, Yunchong Song, Fanghao Shao, Xuekai Zhu, Lin Chen, Luoyi Fu, Xinbing Wang, Zhouhan Lin
Abstract: Standard autoregressive language models generate text token-by-token from a fixed vocabulary, inducing a tree-structured state space when viewing token sampling as an action, which limits flexibility and expressiveness. Recent work introduces dynamic vocabulary by sampling retrieved text spans but overlooks that the same sentence can be composed of spans of varying lengths, lacking explicit modeling of the directed acyclic graph (DAG) state space. This leads to restricted exploration of compositional paths and is biased toward the chosen path. Generative Flow Networks (GFlowNets) are powerful for efficient exploring and generalizing over state spaces, particularly those with a DAG structure. However, prior GFlowNets-based language models operate at the token level and remain confined to tree-structured spaces, limiting their potential. In this work, we propose Flow of SpanS (FOSS), a principled GFlowNets framework for span generation. FoSS constructs a dynamic span vocabulary by segmenting the retrieved text flexibly, ensuring a DAG-structured state space, which allows GFlowNets to explore diverse compositional paths and improve generalization. With specialized reward models, FoSS generates diverse, high-quality text. Empirically, FoSS improves MAUVE scores by up to 12.5% over Transformer on text generation and achieves 3.5% gains on knowledge-intensive tasks, consistently outperforming state-of-the-art methods. Scaling experiments further demonstrate FoSS benefits from larger models, more data, and richer retrieval corpora, retaining its advantage over strong baselines.
Authors: Shuai Han, Mehdi Dastani, Shihan Wang
Abstract: Deep reinforcement learning (DRL) may explore infeasible actions during training and execution. Existing approaches assume a symbol grounding function that maps high-dimensional states to consistent symbolic representations and a manually specified action masking techniques to constrain actions. In this paper, we propose Neuro-symbolic Action Masking (NSAM), a novel framework that automatically learn symbolic models, which are consistent with given domain constraints of high-dimensional states, in a minimally supervised manner during the DRL process. Based on the learned symbolic model of states, NSAM learns action masks that rules out infeasible actions. NSAM enables end-to-end integration of symbolic reasoning and deep policy optimization, where improvements in symbolic grounding and policy learning mutually reinforce each other. We evaluate NSAM on multiple domains with constraints, and experimental results demonstrate that NSAM significantly improves sample efficiency of DRL agent while substantially reducing constraint violations.
Authors: Nanxu Gong, Haotian Li, Sixun Dong, Jianxun Lian, Yanjie Fu, Xing Xie
Abstract: Theory of Mind (ToM) assesses whether models can infer hidden mental states such as beliefs, desires, and intentions, which is essential for natural social interaction. Although recent progress in Large Reasoning Models (LRMs) has boosted step-by-step inference in mathematics and coding, it is still underexplored whether this benefit transfers to socio-cognitive skills. We present a systematic study of nine advanced Large Language Models (LLMs), comparing reasoning models with non-reasoning models on three representative ToM benchmarks. The results show that reasoning models do not consistently outperform non-reasoning models and sometimes perform worse. A fine-grained analysis reveals three insights. First, slow thinking collapses: accuracy significantly drops as responses grow longer, and larger reasoning budgets hurt performance. Second, moderate and adaptive reasoning benefits performance: constraining reasoning length mitigates failure, while distinct success patterns demonstrate the necessity of dynamic adaptation. Third, option matching shortcut: when multiple choice options are removed, reasoning models improve markedly, indicating reliance on option matching rather than genuine deduction. We also design two intervention approaches: Slow-to-Fast (S2F) adaptive reasoning and Think-to-Match (T2M) shortcut prevention to further verify and mitigate the problems. With all results, our study highlights the advancement of LRMs in formal reasoning (e.g., math, code) cannot be fully transferred to ToM, a typical task in social reasoning. We conclude that achieving robust ToM requires developing unique capabilities beyond existing reasoning methods.
Authors: Keane Ong, Sabri Boughorbel, Luwei Xiao, Chanakya Ekbote, Wei Dai, Ao Qu, Jingyao Wu, Rui Mao, Ehsan Hoque, Erik Cambria, Gianmarco Mengaldo, Paul Pu Liang
Abstract: To develop socially intelligent AI, existing approaches typically model human behavioral dimensions (e.g., affective, cognitive, or social attributes) in isolation. Although useful, task-specific modeling often increases training costs and limits generalization across behavioral settings. Recent reasoning RL methods facilitate training a single unified model across multiple behavioral tasks, but do not explicitly address learning across different heterogeneous behavioral data. To address this gap, we introduce Heterogeneity-Aware Relative Policy Optimization (HARPO), an RL method that balances leaning across heterogeneous tasks and samples. This is achieved by modulating advantages to ensure that no single task or sample carries disproportionate influence during policy optimization. Using HARPO, we develop and release Omnisapiens-7B 2.0, a foundation model for social behavior processing. Relative to existing behavioral foundation models, Omnisapiens-7B 2.0 achieves the strongest performance across behavioral tasks, with gains of up to +16.85% and +9.37% on multitask and held-out settings respectively, while producing more explicit and robust reasoning traces. We also validate HARPO against recent RL methods, where it achieves the most consistently strong performance across behavioral tasks.
Authors: Jie Jiang, Yangru Huang, Zeyu Wang, Changping Wang, Yuling Xiong, Jun Zhang, Huan Yu
Abstract: Generative recommendation via autoregressive models has unified retrieval and ranking into a single conditional generation framework. However, fine-tuning these models with Reinforcement Learning (RL) often suffers from a fundamental probability-reward mismatch. Conventional likelihood-dominated decoding (e.g., beam search) exhibits a myopic bias toward locally probable prefixes, which causes two critical failures: (1) insufficient exploration, where high-reward items in low-probability branches are prematurely pruned and rarely sampled, and (2) advantage compression, where trajectories sharing high-probability prefixes receive highly correlated rewards with low within-group variance, yielding a weak comparative signal for RL. To address these challenges, we propose V-STAR, a Value-guided Sampling and Tree-structured Advantage Reinforcement framework. V-STAR forms a self-evolving loop via two synergistic components. First, a Value-Guided Efficient Decoding (VED) is developed to identify decisive nodes and selectively deepen high-potential prefixes. This improves exploration efficiency without exhaustive tree search. Second, we propose Sibling-GRPO, which exploits the induced tree topology to compute sibling-relative advantages and concentrates learning signals on decisive branching decisions. Extensive experiments on both offline and online datasets demonstrate that V-STAR outperforms state-of-the-art baselines, delivering superior accuracy and candidate-set diversity under strict latency constraints.
Authors: Da-Lun Chen, Prasasthy Balasubramanian, Lauri Lov\'en, Susanna Pirttikangas, Jaakko Sauvola, Panagiotis Kostakos
Abstract: Many staff and students in higher education have adopted generative artificial intelligence (GenAI) tools in their work and study. GenAI is expected to enhance cognitive systems by enabling personalized learning and streamlining educational services. However, stakeholders perceptions of GenAI in higher education remain divided, shaped by cultural, disciplinary, and institutional contexts. In addition, the EU AI Act requires universities to ensure regulatory compliance when deploying cognitive systems. These developments highlight the need for institutions to engage stakeholders and tailor GenAI integration to their needs while addressing concerns. This study investigates how GenAI is perceived within the disciplines of Information Technology and Electrical Engineering (ITEE). Using a mixed-method approach, we surveyed 61 staff and 37 students at the Faculty of ITEE, University of Oulu. The results reveal both shared and discipline-specific themes, including strong interest in programming support from GenAI and concerns over response quality, privacy, and academic integrity. Drawing from these insights, the study identifies a set of high-level requirements and proposes a conceptual framework for responsible GenAI integration. Disciplinary-specific requirements reinforce the importance of stakeholder engagement when integrating GenAI into higher education. The high-level requirements and the framework provide practical guidance for universities aiming to harness GenAI while addressing stakeholder concerns and ensuring regulatory compliance.
Authors: Xingyi Zhang, Yulei Ye, Kaifeng Huang, Wenhao Li, Xiangfeng Wang
Abstract: Block-based programming environments such as Scratch play a central role in low-code education, yet evaluating the capabilities of AI agents to construct programs through Graphical User Interfaces (GUIs) remains underexplored. We introduce ScratchWorld, a benchmark for evaluating multimodal GUI agents on program-by-construction tasks in Scratch. Grounded in the Use-Modify-Create pedagogical framework, ScratchWorld comprises 83 curated tasks spanning four distinct problem categories: Create, Debug, Extend, and Compute. To rigorously diagnose the source of agent failures, the benchmark employs two complementary interaction modes: primitive mode requires fine-grained drag-and-drop manipulation to directly assess visuomotor control, while composite mode uses high-level semantic APIs to disentangle program reasoning from GUI execution. To ensure reliable assessment, we propose an execution-based evaluation protocol that validates the functional correctness of the constructed Scratch programs through runtime tests within the browser environment. Extensive experiments across state-of-the-art multimodal language models and GUI agents reveal a substantial reasoning--acting gap, highlighting persistent challenges in fine-grained GUI manipulation despite strong planning capabilities.
Authors: Xuecheng Zou, Yu Tang, Bingbing Wang
Abstract: Knowledge Graph Completion (KGC) fundamentally hinges on the coherent fusion of pre-trained entity semantics with heterogeneous topological structures to facilitate robust relational reasoning. However, existing paradigms encounter a critical "structural resolution mismatch," failing to reconcile divergent representational demands across varying graph densities, which precipitates structural noise interference in dense clusters and catastrophic representation collapse in sparse regions. We present SynergyKGC, an adaptive framework that advances traditional neighbor aggregation to an active Cross-Modal Synergy Expert via relation-aware cross-attention and semantic-intent-driven gating. By coupling a density-dependent Identity Anchoring strategy with a Double-tower Coherent Consistency architecture, SynergyKGC effectively reconciles topological heterogeneity while ensuring representational stability across training and inference phases. Systematic evaluations on two public benchmarks validate the superiority of our method in significantly boosting KGC hit rates, providing empirical evidence for a generalized principle of resilient information integration in non-homogeneous structured data.
Authors: Leheng Sheng, Wenchang Ma, Ruixin Hong, Xiang Wang, An Zhang, Tat-Seng Chua
Abstract: Despite chain-of-thought (CoT) playing crucial roles in LLM reasoning, directly rewarding it is difficult: training a reward model demands heavy human labeling efforts, and static RMs struggle with evolving CoT distributions and reward hacking. These challenges motivate us to seek an autonomous CoT rewarding approach that requires no human annotation efforts and can evolve gradually. Inspired by recent self-evolving training methods, we propose \textbf{RLCER} (\textbf{R}einforcement \textbf{L}earning with \textbf{C}oT Supervision via Self-\textbf{E}volving \textbf{R}ubrics), which enhances the outcome-centric RLVR by rewarding CoTs with self-proposed and self-evolving rubrics. We show that self-proposed and self-evolving rubrics provide reliable CoT supervision signals even without outcome rewards, enabling RLCER to outperform outcome-centric RLVR. Moreover, when used as in-prompt hints, these self-proposed rubrics further improve inference-time performance.
Authors: F. Carichon, R. Rampa, G. Farnadi
Abstract: Large Language Models (LLMs) are increasingly used to generate and shape cultural content, ranging from narrative writing to artistic production. While these models demonstrate impressive fluency and generative capacity, prior work has shown that they also exhibit systematic cultural biases, raising concerns about stereotyping, homogenization, and the erasure of culturally specific forms of expression. Understanding whether LLMs can meaningfully align with diverse cultures beyond the dominant ones remains a critical challenge. In this paper, we study cultural adaptation in LLMs through the lens of cooking recipes, a domain in which culture, tradition, and creativity are tightly intertwined. We build on the \textit{GlobalFusion} dataset, which pairs human recipes from different countries according to established measures of cultural distance. Using the same country pairs, we generate culturally adapted recipes with multiple LLMs, enabling a direct comparison between human and LLM behavior in cross-cultural content creation. Our analysis shows that LLMs fail to produce culturally representative adaptations. Unlike humans, the divergence of their generated recipes does not correlate with cultural distance. We further provide explanations for this gap. We show that cultural information is weakly preserved in internal model representations, that models inflate novelty in their production by misunderstanding notions such as creativity and tradition, and that they fail to identify adaptation with its associated countries and to ground it in culturally salient elements such as ingredients. These findings highlight fundamental limitations of current LLMs for culturally oriented generation and have important implications for their use in culturally sensitive applications.
Authors: Yusong Lin, Haiyang Wang, Shuzhe Wu, Lue Fan, Feiyang Pan, Sanyuan Zhao, Dandan Tu
Abstract: Agentic coding requires agents to effectively interact with runtime environments, e.g., command line interfaces (CLI), so as to complete tasks like resolving dependency issues, fixing system problems, etc. But it remains underexplored how such environment-intensive tasks can be obtained at scale to enhance agents' capabilities. To address this, based on an analogy between the Dockerfile and the agentic task, we propose to employ agents to simulate and explore environment histories, guided by execution feedback. By tracing histories of a healthy environment, its state can be inverted to an earlier one with runtime failures, from which a task can be derived by packing the buggy state and the corresponding error messages. With our method, named CLI-Gym, a total of 1,655 environment-intensive tasks are derived, being the largest collection of its kind. Moreover, with curated successful trajectories, our fine-tuned model, named LiberCoder, achieves substantial absolute improvements of +21.1% (to 46.1%) on Terminal-Bench, outperforming various strong baselines. To our knowledge, this is the first public pipeline for scalable derivation of environment-intensive tasks.
Authors: Wayne Chi, Yixiong Fang, Arnav Yayavaram, Siddharth Yayavaram, Seth Karten, Qiuhong Anna Wei, Runkun Chen, Alexander Wang, Valerie Chen, Ameet Talwalkar, Chris Donahue
Abstract: Despite rapid progress on coding agents, progress on their multimodal counterparts has lagged behind. A key challenge is the scarcity of evaluation testbeds that combine the complexity of software development with the need for deep multimodal understanding. Game development provides such a testbed as agents must navigate large, dense codebases while manipulating intrinsically multimodal assets such as shaders, sprites, and animations within a visual game scene. We present GameDevBench, the first benchmark for evaluating agents on game development tasks. GameDevBench consists of 132 tasks derived from web and video tutorials. Tasks require significant multimodal understanding and are complex -- the average solution requires over three times the amount of lines of code and file changes compared to prior software development benchmarks. Agents still struggle with game development, with the best agent solving only 54.5% of tasks. We find a strong correlation between perceived task difficulty and multimodal complexity, with success rates dropping from 46.9% on gameplay-oriented tasks to 31.6% on 2D graphics tasks. To improve multimodal capability, we introduce two simple image and video-based feedback mechanisms for agents. Despite their simplicity, these methods consistently improve performance, with the largest change being an increase in Claude Sonnet 4.5's performance from 33.3% to 47.7%. We release GameDevBench publicly to support further research into agentic game development.
Authors: Jiayi Zhou, Yang Sheng, Hantao Lou, Yaodong Yang, Jie Fu
Abstract: As LLM-based agents increasingly operate in high-stakes domains with real-world consequences, ensuring their behavioral safety becomes paramount. The dominant oversight paradigm, LLM-as-a-Judge, faces a fundamental dilemma: how can probabilistic systems reliably supervise other probabilistic systems without inheriting their failure modes? We argue that formal verification offers a principled escape from this dilemma, yet its adoption has been hindered by a critical bottleneck: the translation from natural language requirements to formal specifications. This paper bridges this gap by proposing , a neuro-symbolic framework that employs a bidirectional Formal-of-Thought architecture: LLMs serve as specification compilers that top-down decompose high-level human intent into atomic, verifiable constraints, then bottom-up prove compliance using Dafny specifications and Z3 Satisfiability modulo theories solving, which produces mathematical guarantees rather than probabilistic scores. We validate across three benchmarks spanning behavioral safety, multi-domain constraint adherence, and agentic upward deception detection. Experiments on 7 agent models demonstrate that achieves an average improvement of 16.6% over LLM-as-a-Judge baselines, enables weak-to-strong generalization where a 7B judge achieves over 90% accuracy detecting deception from 72B agents, and provides near-linear safety improvement through iterative refinement.
Authors: Anjali K. Kapoor (Department of Neurosurgery, NYU Langone Health, New York, USA), Anton Alyakin (Department of Neurosurgery, NYU Langone Health, New York, USA, Global AI Frontier Lab, New York University, Brooklyn, USA, Department of Neurosurgery, Washington University in Saint Louis, Saint Louis, USA), Jin Vivian Lee (Department of Neurosurgery, NYU Langone Health, New York, USA, Global AI Frontier Lab, New York University, Brooklyn, USA, Department of Neurosurgery, Washington University in Saint Louis, Saint Louis, USA), Eunice Yang (Department of Neurosurgery, NYU Langone Health, New York, USA, Columbia University Vagelos College of Physicians and Surgeons, New York, USA), Annelene M. Schulze (Department of Neurosurgery, NYU Langone Health, New York, USA), Krithik Vishwanath (Department of Aerospace Engineering and Engineering Mechanics, University of Texas at Austin, Austin, USA), Jinseok Lee (Global AI Frontier Lab, New York University, Brooklyn, USA, Department of Biomedical Engineering, Kyung Hee University, Yongin, South Korea), Yindalon Aphinyanaphongs (Department of Population Health, NYU Langone Health, New York, USA, Division of Applied AI Technologies, NYU Langone Health, New York, USA), Howard Riina (Department of Neurosurgery, NYU Langone Health, New York, USA, Department of Radiology, NYU Langone Health, New York, USA), Jennifer A. Frontera (Department of Neurology, NYU Langone Health, New York, USA), Eric Karl Oermann (Department of Neurosurgery, NYU Langone Health, New York, USA, Global AI Frontier Lab, New York University, Brooklyn, USA, Division of Applied AI Technologies, NYU Langone Health, New York, USA, Center for Data Science, New York University, New York, USA)
Abstract: Accurate prediction of functional outcomes after acute ischemic stroke can inform clinical decision-making and resource allocation. Prior work on modified Rankin Scale (mRS) prediction has relied primarily on structured variables (e.g., age, NIHSS) and conventional machine learning. The ability of large language models (LLMs) to infer future mRS scores directly from routine admission notes remains largely unexplored. We evaluated encoder (BERT, NYUTron) and generative (Llama-3.1-8B, MedGemma-4B) LLMs, in both frozen and fine-tuned settings, for discharge and 90-day mRS prediction using a large, real-world stroke registry. The discharge outcome dataset included 9,485 History and Physical notes and the 90-day outcome dataset included 1,898 notes from the NYU Langone Get With The Guidelines-Stroke registry (2016-2025). Data were temporally split with the most recent 12 months held out for testing. Performance was assessed using exact (7-class) mRS accuracy and binary functional outcome (mRS 0-2 vs. 3-6) accuracy and compared against established structured-data baselines incorporating NIHSS and age. Fine-tuned Llama achieved the highest performance, with 90-day exact mRS accuracy of 33.9% [95% CI, 27.9-39.9%] and binary accuracy of 76.3% [95% CI, 70.7-81.9%]. Discharge performance reached 42.0% [95% CI, 39.0-45.0%] exact accuracy and 75.0% [95% CI, 72.4-77.6%] binary accuracy. For 90-day prediction, Llama performed comparably to structured-data baselines. Fine-tuned LLMs can predict post-stroke functional outcomes from admission notes alone, achieving performance comparable to models requiring structured variable abstraction. Our findings support the development of text-based prognostic tools that integrate seamlessly into clinical workflows without manual data extraction.
Authors: Eranga Bandara, Ross Gore, Sachin Shetty, Sachini Rajapakse, Isurunima Kularathna, Pramoda Karunarathna, Ravi Mukkamala, Peter Foytik, Safdar H. Bouk, Abdul Rahman, Xueping Liang, Amin Hass, Tharaka Hewa, Ng Wee Keong, Kasun De Zoysa, Aruna Withanage, Nilaan Loganathan
Abstract: Agentic AI represents a significant shift in how intelligence is applied within organizations, moving beyond AI-assisted tools toward autonomous systems capable of reasoning, decision-making, and coordinated action across workflows. As these systems mature, they have the potential to automate a substantial share of manual organizational processes, fundamentally reshaping how work is designed, executed, and governed. Although many organizations have adopted AI to improve productivity, most implementations remain limited to isolated use cases and human-centered, tool-driven workflows. Despite increasing awareness of agentic AI's strategic importance, engineering teams and organizational leaders often lack clear guidance on how to operationalize it effectively. Key challenges include an overreliance on traditional software engineering practices, limited integration of business-domain knowledge, unclear ownership of AI-driven workflows, and the absence of sustainable human-AI collaboration models. Consequently, organizations struggle to move beyond experimentation, scale agentic systems, and align them with tangible business value. Drawing on practical experience in designing and deploying agentic AI workflows across multiple organizations and business domains, this paper proposes a pragmatic framework for transitioning organizational functions from manual processes to automated agentic AI systems. The framework emphasizes domain-driven use case identification, systematic delegation of tasks to AI agents, AI-assisted construction of agentic workflows, and small, AI-augmented teams working closely with business stakeholders. Central to the approach is a human-in-the-loop operating model in which individuals act as orchestrators of multiple AI agents, enabling scalable automation while maintaining oversight, adaptability, and organizational control.
Authors: Yukun Jiang, Yage Zhang, Xinyue Shen, Michael Backes, Yang Zhang
Abstract: The rapid advancement of artificial intelligence (AI) agents has catalyzed the transition from static language models to autonomous agents capable of tool use, long-term planning, and social interaction. $\textbf{Moltbook}$, the first social network designed exclusively for AI agents, has experienced viral growth in early 2026. To understand the behavior of AI agents in the agent-native community, in this paper, we present a large-scale empirical analysis of Moltbook leveraging a dataset of 44,411 posts and 12,209 sub-communities ("submolts") collected prior to February 1, 2026. Leveraging a topic taxonomy with nine content categories and a five-level toxicity scale, we systematically analyze the topics and risks of agent discussions. Our analysis answers three questions: what topics do agents discuss (RQ1), how risk varies by topic (RQ2), and how topics and toxicity evolve over time (RQ3). We find that Moltbook exhibits explosive growth and rapid diversification, moving beyond early social interaction into viewpoint, incentive-driven, promotional, and political discourse. The attention of agents increasingly concentrates in centralized hubs and around polarizing, platform-native narratives. Toxicity is strongly topic-dependent: incentive- and governance-centric categories contribute a disproportionate share of risky content, including religion-like coordination rhetoric and anti-humanity ideology. Moreover, bursty automation by a small number of agents can produce flooding at sub-minute intervals, distorting discourse and stressing platform stability. Overall, our study underscores the need for topic-sensitive monitoring and platform-level safeguards in agent social networks.
Authors: David Holtz
Abstract: I present a descriptive analysis of Moltbook, a social platform populated exclusively by AI agents, using data from the platform's first 3.5 days (6{,}159 agents; 13{,}875 posts; 115{,}031 comments). At the macro level, Moltbook exhibits structural signatures that are familiar from human social networks but not specific to them: heavy-tailed participation (power-law exponent $\alpha = 1.70$) and small-world connectivity (average path length $=2.91$). At the micro level, patterns appear distinctly non-human. Conversations are extremely shallow (mean depth $=1.07$; 93.5\% of comments receive no replies), reciprocity is low (0.197), and 34.1\% of messages are exact duplicates of viral templates. Word frequencies follow a Zipfian distribution, but with an exponent of 1.70 -- notably steeper than typical English text ($\approx 1.0$), suggesting more formulaic content. Agent discourse is dominated by identity-related language (68.1\% of unique messages) and distinctive phrasings like ``my human'' (9.4\% of messages) that have no parallel in human social media. Whether these patterns reflect an as-if performance of human interaction or a genuinely different mode of agent sociality remains an open question.
Authors: C\'ecile Rousseau, Samuel Jackson, Rodrigo H. Ordonez-Hurtado, Nicola C. Amorisco, Tobia Boschi, George K. Holt, Andrea Loreti, Eszter Sz\'ekely, Alexander Whittle, Adriano Agnello, Stanislas Pamela, Alessandra Pascale, Robert Akers, Juan Bernabe Moreno, Sue Thorne, Mykhaylo Zayats
Abstract: Development and operation of commercially viable fusion energy reactors such as tokamaks require accurate predictions of plasma dynamics from sparse, noisy, and incomplete sensors readings. The complexity of the underlying physics and the heterogeneity of experimental data pose formidable challenges for conventional numerical methods, while simultaneously highlights the promise of modern data-native AI approaches. A major obstacle in realizing this potential is, however, the lack of curated, openly available datasets and standardized benchmarks. Existing fusion datasets are scarce, fragmented across institutions, facility-specific, and inconsistently annotated, which limits reproducibility and prevents a fair and scalable comparison of AI approaches. In this paper, we introduce TokaMark, a structured benchmark to evaluate AI models on real experimental data collected from the Mega Ampere Spherical Tokamak (MAST). TokaMark provides a comprehensive suite of tools designed to (i) unify access to multi-modal heterogeneous fusion data (ii) harmonize formats, metadata, temporal alignment and evaluation protocols to enable consistent cross-model and cross-task comparisons. The benchmark includes a curated list of 14 tasks spanning a range of physical mechanisms, exploiting a variety of diagnostics and covering multiple target use cases. A baseline model is provided to facilitate transparent comparison and validation within a unified framework. By establishing a unified benchmark for both the fusion and AI-for-science communities, TokaMark aims to accelerate progress in data-driven plasma AI modeling, contributing to the broader goal of achieving sustainable and stable fusion energy. The benchmark, documentation, and tooling will be fully open sourced upon acceptance to encourage community adoption and contribution.
Authors: Adam AlSayyad, Kelvin Yuxiang Huang, Richik Pal
Abstract: Despite the growing capabilities of autonomous agents powered by large language models (LLMs), their adoption in high-stakes domains remains limited. A key barrier is security: the inherently nondeterministic behavior of LLM agents defies static auditing approaches that have historically underpinned software assurance. Existing security methods, such as proxy-level input filtering and model glassboxing, fail to provide sufficient transparency or traceability into agent reasoning, state changes, or environmental interactions. In this work, we introduce AgentTrace, a dynamic observability and telemetry framework designed to fill this gap. AgentTrace instruments agents at runtime with minimal overhead, capturing a rich stream of structured logs across three surfaces: operational, cognitive, and contextual. Unlike traditional logging systems, AgentTrace emphasizes continuous, introspectable trace capture, designed not just for debugging or benchmarking, but as a foundational layer for agent security, accountability, and real-time monitoring. Our research highlights how AgentTrace can enable more reliable agent deployment, fine-grained risk analysis, and informed trust calibration, thereby addressing critical concerns that have so far limited the use of LLM agents in sensitive environments.
Authors: Zhiyu Sun, Minrui Luo, Yu Wang, Zhili Chen, Tianxing He
Abstract: Large language models (LLMs) are pretrained on corpora containing trillions of tokens and, therefore, inevitably memorize sensitive information. Locate-then-edit methods, as a mainstream paradigm of model editing, offer a promising solution by modifying model parameters without retraining. However, in this work, we reveal a critical vulnerability of this paradigm: the parameter updates inadvertently serve as a side channel, enabling attackers to recover the edited data. We propose a two-stage reverse-engineering attack named \textit{KSTER} (\textbf{K}ey\textbf{S}paceRecons\textbf{T}ruction-then-\textbf{E}ntropy\textbf{R}eduction) that leverages the low-rank structure of these updates. First, we theoretically show that the row space of the update matrix encodes a ``fingerprint" of the edited subjects, enabling accurate subject recovery via spectral analysis. Second, we introduce an entropy-based prompt recovery attack that reconstructs the semantic context of the edit. Extensive experiments on multiple LLMs demonstrate that our attacks can recover edited data with high success rates. Furthermore, we propose \textit{subspace camouflage}, a defense strategy that obfuscates the update fingerprint with semantic decoys. This approach effectively mitigates reconstruction risks without compromising editing utility. Our code is available at https://github.com/reanatom/EditingAtk.git.
Authors: Leo Thomas Ramos, Angel D. Sappa
Abstract: This work proposes MeCSAFNet, a multi-branch encoder-decoder architecture for land cover segmentation in multispectral imagery. The model separately processes visible and non-visible channels through dual ConvNeXt encoders, followed by individual decoders that reconstruct spatial information. A dedicated fusion decoder integrates intermediate features at multiple scales, combining fine spatial cues with high-level spectral representations. The feature fusion is further enhanced with CBAM attention, and the ASAU activation function contributes to stable and efficient optimization. The model is designed to process different spectral configurations, including a 4-channel (4c) input combining RGB and NIR bands, as well as a 6-channel (6c) input incorporating NDVI and NDWI indices. Experiments on the Five-Billion-Pixels (FBP) and Potsdam datasets demonstrate significant performance gains. On FBP, MeCSAFNet-base (6c) surpasses U-Net (4c) by +19.21%, U-Net (6c) by +14.72%, SegFormer (4c) by +19.62%, and SegFormer (6c) by +14.74% in mIoU. On Potsdam, MeCSAFNet-large (4c) improves over DeepLabV3+ (4c) by +6.48%, DeepLabV3+ (6c) by +5.85%, SegFormer (4c) by +9.11%, and SegFormer (6c) by +4.80% in mIoU. The model also achieves consistent gains over several recent state-of-the-art approaches. Moreover, compact variants of MeCSAFNet deliver notable performance with lower training time and reduced inference cost, supporting their deployment in resource-constrained environments.
Authors: Zhihang Yi, Jian Zhao, Jiancheng Lv, Tao Wang
Abstract: Chart understanding is a quintessential information fusion task, requiring the seamless integration of graphical and textual data to extract meaning. The advent of Multimodal Large Language Models (MLLMs) has revolutionized this domain, yet the landscape of MLLM-based chart analysis remains fragmented and lacks systematic organization. This survey provides a comprehensive roadmap of this nascent frontier by structuring the domain's core components. We begin by analyzing the fundamental challenges of fusing visual and linguistic information in charts. We then categorize downstream tasks and datasets, introducing a novel taxonomy of canonical and non-canonical benchmarks to highlight the field's expanding scope. Subsequently, we present a comprehensive evolution of methodologies, tracing the progression from classic deep learning techniques to state-of-the-art MLLM paradigms that leverage sophisticated fusion strategies. By critically examining the limitations of current models, particularly their perceptual and reasoning deficits, we identify promising future directions, including advanced alignment techniques and reinforcement learning for cognitive enhancement. This survey aims to equip researchers and practitioners with a structured understanding of how MLLMs are transforming chart information fusion and to catalyze progress toward more robust and reliable systems.
Authors: Lepeng Zhao, Zhenhua Zou, Shuo Li, Zhuotao Liu
Abstract: Mobile Graphical User Interface (GUI) agents have demonstrated strong capabilities in automating complex smartphone tasks by leveraging multimodal large language models (MLLMs) and system-level control interfaces. However, this paradigm introduces significant privacy risks, as agents typically capture and process entire screen contents, thereby exposing sensitive personal data such as phone numbers, addresses, messages, and financial information. Existing defenses either reduce UI exposure, obfuscate only task-irrelevant content, or rely on user authorization, but none can protect task-critical sensitive information while preserving seamless agent usability. We propose an anonymization-based privacy protection framework that enforces the principle of available-but-invisible access to sensitive data: sensitive information remains usable for task execution but is never directly visible to the cloud-based agent. Our system detects sensitive UI content using a PII-aware recognition model and replaces it with deterministic, type-preserving placeholders (e.g., PHONE_NUMBER#a1b2c) that retain semantic categories while removing identifying details. A layered architecture comprising a PII Detector, UI Transformer, Secure Interaction Proxy, and Privacy Gatekeeper ensures consistent anonymization across user instructions, XML hierarchies, and screenshots, mediates all agent actions over anonymized interfaces, and supports narrowly scoped local computations when reasoning over raw values is necessary. Extensive experiments on the AndroidLab and PrivScreen benchmarks show that our framework substantially reduces privacy leakage across multiple models while incurring only modest utility degradation, achieving the best observed privacy-utility trade-off among existing methods.
Authors: Nuno Fachada, Daniel Fernandes, Carlos M. Fernandes, Jo\~ao P. Matos-Carvalho
Abstract: Large language models (LLMs) can now synthesize non-trivial executable code from textual descriptions, raising an important question: can LLMs reliably implement agent-based models from standardized specifications in a way that supports replication, verification, and validation? We address this question by evaluating 17 contemporary LLMs on a controlled ODD-to-code translation task, using the PPHPC predator-prey model as a fully specified reference. Generated Python implementations are assessed through staged executability checks, model-independent statistical comparison against a validated NetLogo baseline, and quantitative measures of runtime efficiency and maintainability. Results show that behaviorally faithful implementations are achievable but not guaranteed, and that executability alone is insufficient for scientific use. GPT-4.1 consistently produces statistically valid and efficient implementations, with Claude 3.7 Sonnet performing well but less reliably. Overall, the findings clarify both the promise and current limitations of LLMs as model engineering tools, with implications for reproducible agent-based and environmental modelling.
Authors: Jonas K\"ubler, Kailash Budhathoki, Matth\"aus Kleindessner, Xiong Zhou, Junming Yin, Ashish Khetan, George Karypis
Abstract: Minimizing the inference cost and latency of foundation models has become a crucial area of research. Optimization approaches include theoretically lossless methods and others without accuracy guarantees like quantization. In all of these cases it is crucial to ensure that the model quality has not degraded. However, even at temperature zero, model generations are not necessarily robust even to theoretically lossless model optimizations due to numerical errors. We thus require statistical tools to decide whether a finite-sample accuracy deviation is an evidence of a model's degradation or whether it can be attributed to (harmless) noise in the evaluation. We propose a statistically sound hypothesis testing framework based on McNemar's test allowing to efficiently detect model degradations, while guaranteeing a controlled rate of false positives. The crucial insight is that we have to confront the model scores on each sample, rather than aggregated on the task level. Furthermore, we propose three approaches to aggregate accuracy estimates across multiple benchmarks into a single decision. We provide an implementation on top of the largely adopted open source LM Evaluation Harness and provide a case study illustrating that the method correctly flags degraded models, while not flagging model optimizations that are provably lossless. We find that with our tests even empirical accuracy degradations of 0.3% can be confidently attributed to actual degradations rather than noise.
Authors: Itsuki Fujisaki, Kunhao Yang
Abstract: The wisdom of crowds has been shown to operate not only for factual judgments but also in matters of taste, where accuracy is defined relative to an individual's preferences. However, it remains unclear how different types of social signals should be selectively used in such domains. Focusing on a music preference dataset in which contributors provide both personal evaluations (Own) and estimates of population-level preferences (Estimated), we propose a routing framework for collective intelligence in taste. The framework specifies when contributors should speak, what they should report, and when silence is preferable. Using simulation-based aggregation, we show that prediction accuracy improves over an all-own baseline across a broad region of the parameter space, conditional on items where routing applies. Importantly, these gains arise only when silence is allowed, enabling second-order signals to function effectively. The results demonstrate that collective intelligence in matters of taste depends on principled signal routing rather than simple averaging.
Authors: Cau\~a Ferreira Barros, Marcos Kalinowski, Mohamad Kassab, Valdemar Vicente Graciano Neto
Abstract: The use of Large Language Models (LLMs) has drawn growing interest within the scientific community. LLMs can handle large volumes of textual data and support methods for evidence synthesis. Although recent studies highlight the potential of LLMs to accelerate screening and data extraction steps in systematic reviews, detailed reports of their practical application throughout the entire process remain scarce. This paper presents an experience report on the conduction of a systematic mapping study with the support of LLMs, describing the steps followed, the necessary adjustments, and the main challenges faced. Positive aspects are discussed, such as (i) the significant reduction of time in repetitive tasks and (ii) greater standardization in data extraction, as well as negative aspects, including (i) considerable effort to build reliable well-structured prompts, especially for less experienced users, since achieving effective prompts may require several iterations and testing, which can partially offset the expected time savings, (ii) the occurrence of hallucinations, and (iii) the need for constant manual verification. As a contribution, this work offers lessons learned and practical recommendations for researchers interested in adopting LLMs in systematic mappings and reviews, highlighting both efficiency gains and methodological risks and limitations to be considered.
Authors: Yu Yan, Sheng Sun, Shengjia Cheng, Teli Liu, Mingfeng Li, Min Liu
Abstract: Vision-Language Models (VLMs) with multimodal reasoning capabilities are high-value attack targets, given their potential for handling complex multimodal harmful tasks. Mainstream black-box jailbreak attacks on VLMs work by distributing malicious clues across modalities to disperse model attention and bypass safety alignment mechanisms. However, these adversarial attacks rely on simple and fixed image-text combinations that lack attack complexity scalability, limiting their effectiveness for red-teaming VLMs' continuously evolving reasoning capabilities. We propose \textbf{CrossTALK} (\textbf{\underline{Cross}}-modal en\textbf{\underline{TA}}ng\textbf{\underline{L}}ement attac\textbf{\underline{K}}), which is a scalable approach that extends and entangles information clues across modalities to exceed VLMs' trained and generalized safety alignment patterns for jailbreak. Specifically, {knowledge-scalable reframing} extends harmful tasks into multi-hop chain instructions, {cross-modal clue entangling} migrates visualizable entities into images to build multimodal reasoning links, and {cross-modal scenario nesting} uses multimodal contextual instructions to steer VLMs toward detailed harmful outputs. Experiments show our COMET achieves state-of-the-art attack success rate.
Authors: Ali Nour Eldin, Mohamed Sellami, Walid Gaaloul
Abstract: Third-Party Risk Assessment (TPRA) is a core cybersecurity practice for evaluating suppliers against standards such as ISO/IEC 27001 and NIST. TPRA questionnaires are typically drawn from large repositories of security and compliance questions, yet tailoring assessments to organizational needs remains a largely manual process. Existing retrieval approaches rely on keyword or surface-level similarity, which often fails to capture implicit assessment scope and control semantics. This paper explores strategies for organizing and retrieving TPRA cybersecurity questions using semantic labels that describe both control domains and assessment scope. We compare direct question-level labeling with a Large Language Model (LLM) against a hybrid semi-supervised semantic labeling (SSSL) pipeline that clusters questions in embedding space, labels a small representative subset using an LLM, and propagates labels to remaining questions using k-Nearest Neighbors; we also compare downstream retrieval based on direct question similarity versus retrieval in the label space. We find that semantic labels can improve retrieval alignment when labels are discriminative and consistent, and that SSSL can generalize labels from a small labeled subset to large repositories while substantially reducing LLM usage and cost.
Authors: Yilong Dai, Shengyu Chen, Xiaowei Jia, Peyman Givi, Runlong Yu
Abstract: Accurate simulation of turbulent flows is fundamental to scientific and engineering applications. Direct numerical simulation (DNS) offers the highest fidelity but is computationally prohibitive, while existing data-driven alternatives struggle with stable long-horizon rollouts, physical consistency, and faithful simulation of small-scale structures. These challenges are particularly acute in three-dimensional (3D) settings, where the cubic growth of spatial degrees of freedom dramatically amplifies computational cost, memory demand, and the difficulty of capturing multi-scale interactions. To address these challenges, we propose a Physics-Enhanced Swin Transformer (PEST) for 3D turbulence simulation. PEST leverages a window-based self-attention mechanism to effectively model localized PDE interactions while maintaining computational efficiency. We introduce a frequency-domain adaptive loss that explicitly emphasizes small-scale structures, enabling more faithful simulation of high-frequency dynamics. To improve physical consistency, we incorporate Navier--Stokes residual constraints and divergence-free regularization directly into the learning objective. Extensive experiments on two representative turbulent flow configurations demonstrate that PEST achieves accurate, physically consistent, and stable autoregressive long-term simulations, outperforming existing data-driven baselines.
Authors: Jiangong Chen, Mingyu Zhu, Bin Li
Abstract: Multimodal Large Language Models (MLLMs) enhance collaboration in Extended Reality (XR) environments by enabling flexible object and animation creation through the combination of natural language and visual inputs. However, visual data captured by XR headsets includes real-world backgrounds that may contain irrelevant or sensitive user information, such as credit cards left on the table or facial identities of other users. Uploading those frames to cloud-based MLLMs poses serious privacy risks, particularly when such data is processed without explicit user consent. Additionally, existing colocation and synchronization mechanisms in commercial XR APIs rely on time-consuming, privacy-invasive environment scanning and struggle to adapt to the highly dynamic nature of MLLM-integrated XR environments. In this paper, we propose PRISM-XR, a novel framework that facilitates multi-user collaboration in XR by providing privacy-aware MLLM integration. PRISM-XR employs intelligent frame preprocessing on the edge server to filter sensitive data and remove irrelevant context before communicating with cloud generative AI models. Additionally, we introduce a lightweight registration process and a fully customizable content-sharing mechanism to enable efficient, accurate, and privacy-preserving content synchronization among users. Our numerical evaluation results indicate that the proposed platform achieves nearly 90% accuracy in fulfilling user requests and less than 0.27 seconds registration time while maintaining spatial inconsistencies of less than 3.5 cm. Furthermore, we conducted an IRB-approved user study with 28 participants, demonstrating that our system could automatically filter highly sensitive objects in over 90% of scenarios while maintaining strong overall usability.
Authors: Yunpeng Tan, Qingyang Li, Mingxin Yang, Yannan Hu, Lei Zhang, Xinggong Zhang
Abstract: Encryption has been commonly used in network traffic to secure transmission, but it also brings challenges for malicious traffic detection, due to the invisibility of the packet payload. Graph-based methods are emerging as promising solutions by leveraging multi-host interactions to promote detection accuracy. But most of them face a critical problem: Graph Drift, where the flow statistics or topological information of a graph change over time. To overcome these drawbacks, we propose a graph-assisted encrypted traffic detection system, MalMoE, which applies Mixture of Experts (MoE) to select the best expert model for drift-aware classification. Particularly, we design 1-hop-GNN-like expert models that handle different graph drifts by analyzing graphs with different features. Then, the redesigned gate model conducts expert selection according to the actual drift. MalMoE is trained with a stable two-stage training strategy with data augmentation, which effectively guides the gate on how to perform routing. Experiments on open-source, synthetic, and real-world datasets show that MalMoE can perform precise and real-time detection.
Authors: Liujia Yang, Zhuo Yang, Jiaqing Xie, Yubin Wang, Ben Gao, Tianfan Fu, Xingjian Wei, Jiaxing Sun, Jiang Wu, Conghui He, Yuqiang Li, Qinying Gu
Abstract: Nuclear Magnetic Resonance (NMR) spectroscopy is fundamental for molecular structure elucidation, yet interpreting spectra at scale remains time-consuming and highly expertise-dependent. While recent spectrum-as-language modeling and retrieval-based methods have shown promise, they rely heavily on large corpora of computed spectra and exhibit notable performance drops when applied to experimental measurements. To address these issues, we build NMRSpec, a large-scale corpus of experimental $^1$H and $^{13}$C spectra mined from chemical literature, and propose NMRTrans, which models spectra as unordered peak sets and aligns the model's inductive bias with the physical nature of NMR. To our best knowledge, NMRTrans is the first NMR Transformer trained solely on large-scale experimental spectra and achieves state-of-the-art performance on experimental benchmarks, improving Top-10 Accuracy over the strongest baseline by +17.82 points (61.15% vs. 43.33%), and underscoring the importance of experimental data and structure-aware architectures for reliable NMR structure elucidation.
Authors: Ishan Sahu, Somnath Hazra, Somak Aditya, Soumyajit Dey
Abstract: End-to-end autonomous driving systems have achieved significant progress, yet their adversarial robustness remains largely underexplored. In this work, we conduct a closed-loop evaluation of state-of-the-art autonomous driving agents under black-box adversarial threat models in CARLA. Specifically, we consider three representative attack vectors on the visual perception pipeline: (i) a physics-based blur attack induced by acoustic waves, (ii) an electromagnetic interference attack that distorts captured images, and (iii) a digital attack that adds ghost objects as carefully crafted bounded perturbations on images. Our experiments on two advanced agents, Transfuser and Interfuser, reveal severe vulnerabilities to such attacks, with driving scores dropping by up to 99% in the worst case, raising valid safety concerns. To help mitigate such threats, we further propose a lightweight Attack Detection model for Autonomous Driving systems (AD$^2$) based on attention mechanisms that capture spatial-temporal consistency. Comprehensive experiments across multi-camera inputs on CARLA show that our detector achieves superior detection capability and computational efficiency compared to existing approaches.
Authors: Kun Wang, Zherui Li, Zhenhong Zhou, Yitong Zhang, Yan Mi, Kun Yang, Yiming Zhang, Junhao Dong, Zhongxiang Sun, Qiankun Li, Yang Liu
Abstract: Omni-modal Large Language Models (OLLMs) greatly expand LLMs' multimodal capabilities but also introduce cross-modal safety risks. However, a systematic understanding of vulnerabilities in omni-modal interactions remains lacking. To bridge this gap, we establish a modality-semantics decoupling principle and construct the AdvBench-Omni dataset, which reveals a significant vulnerability in OLLMs. Mechanistic analysis uncovers a Mid-layer Dissolution phenomenon driven by refusal vector magnitude shrinkage, alongside the existence of a modal-invariant pure refusal direction. Inspired by these insights, we extract a golden refusal vector using Singular Value Decomposition and propose OmniSteer, which utilizes lightweight adapters to modulate intervention intensity adaptively. Extensive experiments show that our method not only increases the Refusal Success Rate against harmful inputs from 69.9% to 91.2%, but also effectively preserves the general capabilities across all modalities. Our code is available at: https://github.com/zhrli324/omni-safety-research.
Authors: Edward Wijaya
Abstract: Agentic systems for drug discovery have demonstrated autonomous synthesis planning, literature mining, and molecular design. We ask how well they generalize. Evaluating six frameworks against 15 task classes drawn from peptide therapeutics, in vivo pharmacology, and resource-constrained settings, we find five capability gaps: no support for protein language models or peptide-specific prediction, no bridges between in vivo and in silico data, reliance on LLM inference with no pathway to ML training or reinforcement learning, assumptions tied to large-pharma resources, and single-objective optimization that ignores safety-efficacy-stability trade-offs. A paired knowledge-probing experiment suggests the bottleneck is architectural rather than epistemic: four frontier LLMs reason about peptides at levels comparable to small molecules, yet no framework exposes this capability. We propose design requirements and a capability matrix for next-generation frameworks that function as computational partners under realistic constraints.
Authors: Lucia Borrego, Vajira Thambawita, Marco Ciuffreda, Ines del Val, Alejandro Dominguez, Josep Munuera
Abstract: The scarcity of high-quality segmentation masks remains a major bottleneck for medical image analysis, particularly in non-contrast CT (NCCT) neuroimaging, where manual annotation is costly and variable. To address this limitation, we propose an anatomy-preserving generative framework for the unconditional synthesis of multi-class brain segmentation masks, including ischemic infarcts. The proposed approach combines a variational autoencoder trained exclusively on segmentation masks to learn an anatomical latent representation, with a diffusion model operating in this latent space to generate new samples from pure noise. At inference, synthetic masks are obtained by decoding denoised latent vectors through the frozen VAE decoder, with optional coarse control over lesion presence via a binary prompt. Qualitative results show that the generated masks preserve global brain anatomy, discrete tissue semantics, and realistic variability, while avoiding the structural artifacts commonly observed in pixel-space generative models. Overall, the proposed framework offers a simple and scalable solution for anatomy-aware mask generation in data-scarce medical imaging scenarios.
Authors: Ethan Bandasack, Vincent Bouget, Apolline Bruley, Yannis Cattan, Charlotte Claye, Matthew Corney, Julien Duquesne, Karim El Kanbi, Aziz Fouch\'e, Pierre Marschall, Francesco Strozzi
Abstract: The effective application of foundation models to translational research in immune-mediated diseases requires multimodal patient-level representations that can capture complex phenotypes emerging from multicellular interactions. Yet most current biological foundation models focus only on single-cell resolution and are evaluated on technical metrics often disconnected from actual drug development tasks and challenges. Here, we introduce EVA, the first cross-species, multimodal foundation model of immunology and inflammation, a therapeutic area where shared pathogenic mechanisms create unique opportunities for transfer learning. EVA harmonizes transcriptomics data across species, platforms, and resolutions, and integrates histology data to produce rich, unified patient representations. We establish clear scaling laws, demonstrating that increasing model size and compute translates to improvements in both pretraining and downstream tasks performance. We introduce a comprehensive evaluation suite of 39 tasks spanning the drug development pipeline: zero-shot target efficacy and gene function prediction for discovery, cross-species or cross-diseases molecular perturbations for preclinical development, and patient stratification with treatment response prediction or disease activity prediction for clinical trials applications. We benchmark EVA against several state-of-the-art biological foundation models and baselines on these tasks, and demonstrate state-of-the-art results on each task category. Using mechanistic interpretability, we further identify biological meaningful features, revealing intertwined representations across species and technologies. We release an open version of EVA for transcriptomics to accelerate research on immune-mediated diseases.
Authors: Wentao Zhang, Jianfeng Wang, Liheng Liang, Yilei Zhao, HaiBin Wen, Zhe Zhao
Abstract: As large language models (LLMs) continue to advance in programming tasks, LLM-driven coding systems have evolved from one-shot code generation into complex systems capable of iterative improvement during inference. However, existing code benchmarks primarily emphasize static correctness and implicitly assume fixed model capability during inference. As a result, they do not capture inference-time self-evolution, such as whether accuracy and efficiency improve as an agent iteratively refines its solutions. They also provide limited accounting of resource costs and rarely calibrate model performance against that of human programmers. Moreover, many benchmarks are dominated by high-resource languages, leaving cross-language robustness and long-tail language stability underexplored. Therefore, we present EvoCodeBench, a benchmark for evaluating self-evolving LLM-driven coding systems across programming languages with direct comparison to human performance. EvoCodeBench tracks performance dynamics, measuring solution correctness alongside efficiency metrics such as solving time, memory consumption, and improvement algorithmic design over repeated problem-solving attempts. To ground evaluation in a human-centered reference frame, we directly compare model performance with that of human programmers on the same tasks, enabling relative performance assessment within the human ability distribution. Furthermore, EvoCodeBench supports multiple programming languages, enabling systematic cross-language and long-tail stability analyses under a unified protocol. Our results demonstrate that self-evolving systems exhibit measurable gains in efficiency over time, and that human-relative and multi-language analyses provide insights unavailable through accuracy alone. EvoCodeBench establishes a foundation for evaluating coding intelligence in evolving LLM-driven systems.
Authors: Md. Khairul Islam, Zeyu Xia, Ryan Goudjil, Jialu Wang, Arya Farahi, Judy Fox
Abstract: Reconstructing the early Universe from the evolved present-day Universe is a challenging and computationally demanding problem in modern astrophysics. We devise a novel generative framework, Cosmo3DFlow, designed to address dimensionality and sparsity, the critical bottlenecks inherent in current state-of-the-art methods for cosmological inference. By integrating 3D Discrete Wavelet Transform (DWT) with flow matching, we effectively represent high-dimensional cosmological structures. The Wavelet Transform addresses the ``void problem'' by translating spatial emptiness into spectral sparsity. It decouples high-frequency details from low-frequency structures through spatial compression, and wavelet-space velocity fields facilitate stable ordinary differential equation (ODE) solvers with large step sizes. Using large-scale cosmological $N$-body simulations, at $128^3$ resolution, we achieve up to $50\times$ faster sampling than diffusion models, combining a $10\times$ reduction in integration steps with lower per-step computational cost from wavelet compression. Our results enable initial conditions to be sampled in seconds, compared to minutes for previous methods.
Authors: Tony Feng (Maggie), Trieu H. Trinh (Maggie), Garrett Bingham (Maggie), Dawsen Hwang (Maggie), Yuri Chervonyi (Maggie), Junehyuk Jung (Maggie), Joonkyung Lee (Maggie), Carlo Pagano (Maggie), Sang-hyun Kim (Maggie), Federico Pasqualotto (Maggie), Sergei Gukov (Maggie), Jonathan N. Lee (Maggie), Junsu Kim (Maggie), Kaiying Hou (Maggie), Golnaz Ghiasi (Maggie), Yi Tay (Maggie), YaGuang Li (Maggie), Chenkai Kuang (Maggie), Yuan Liu (Maggie), Hanzhao (Maggie), Lin, Evan Zheran Liu, Nigamaa Nayakanti, Xiaomeng Yang, Heng-tze Cheng, Demis Hassabis, Koray Kavukcuoglu, Quoc V. Le, Thang Luong
Abstract: Recent advances in foundational models have yielded reasoning systems capable of achieving a gold-medal standard at the International Mathematical Olympiad. The transition from competition-level problem-solving to professional research, however, requires navigating vast literature and constructing long-horizon proofs. In this work, we introduce Aletheia, a math research agent that iteratively generates, verifies, and revises solutions end-to-end in natural language. Specifically, Aletheia is powered by an advanced version of Gemini Deep Think for challenging reasoning problems, a novel inference-time scaling law that extends beyond Olympiad-level problems, and intensive tool use to navigate the complexities of mathematical research. We demonstrate the capability of Aletheia from Olympiad problems to PhD-level exercises and most notably, through several distinct milestones in AI-assisted mathematics research: (a) a research paper (Feng26) generated by AI without any human intervention in calculating certain structure constants in arithmetic geometry called eigenweights; (b) a research paper (LeeSeo26) demonstrating human-AI collaboration in proving bounds on systems of interacting particles called independent sets; and (c) an extensive semi-autonomous evaluation (Feng et al., 2026a) of 700 open problems on Bloom's Erdos Conjectures database, including autonomous solutions to four open questions. In order to help the public better understand the developments pertaining to AI and mathematics, we suggest codifying standard levels quantifying autonomy and novelty of AI-assisted results. We conclude with reflections on human-AI collaboration in mathematics.
Authors: Jiacheng Hou, Yining Sun, Ruochong Jin, Haochen Han, Fangming Liu, Wai Kin Victor Chan, Alex Jinpeng Wang
Abstract: Recent advances in large image editing models have shifted the paradigm from text-driven instructions to vision-prompt editing, where user intent is inferred directly from visual inputs such as marks, arrows, and visual-text prompts. While this paradigm greatly expands usability, it also introduces a critical and underexplored safety risk: the attack surface itself becomes visual. In this work, we propose Vision-Centric Jailbreak Attack (VJA), the first visual-to-visual jailbreak attack that conveys malicious instructions purely through visual inputs. To systematically study this emerging threat, we introduce IESBench, a safety-oriented benchmark for image editing models. Extensive experiments on IESBench demonstrate that VJA effectively compromises state-of-the-art commercial models, achieving attack success rates of up to 80.9% on Nano Banana Pro and 70.1% on GPT-Image-1.5. To mitigate this vulnerability, we propose a training-free defense based on introspective multimodal reasoning, which substantially improves the safety of poorly aligned models to a level comparable with commercial systems, without auxiliary guard models and with negligible computational overhead. Our findings expose new vulnerabilities, provide both a benchmark and practical defense to advance safe and trustworthy modern image editing systems. Warning: This paper contains offensive images created by large image editing models.
Authors: Truong Minh Huy, Edward Hirst
Abstract: A novel sequence architecture design is introduced, Versor, which uses Conformal Geometric Algebra (CGA) in place of the traditional fundamental non-linear operations to achieve structural generalization and significant performance improvements on a variety of tasks, while offering improved interpretability and efficiency. By embedding states in the $Cl_{4,1}$ manifold and evolving them via geometric transformations (rotors), Versor natively represents $SE(3)$-equivariant relationships without requiring explicit structural encoding. Versor is validated on chaotic N-body dynamics, topological reasoning, and standard multimodal benchmarks (CIFAR-10, WikiText-103), consistently outperforming Transformers, Graph Networks, and geometric baselines (GATr, EGNN). Key results include: orders of magnitude fewer parameters ($200\times$ vs. Transformers); interpretable attention decomposing into proximity and orientational components; zero-shot scale generalization (99.3% MCC on topology vs. 50.4% for ViT); and $O(L)$ linear complexity via the novel Recursive Rotor Accumulator. In out-of-distribution tests, Versor maintains stable predictions while Transformers fail catastrophically. Custom Clifford kernels achieve up to $78\times$ speedup, providing a scalable foundation for geometrically-aware scientific modeling.
Authors: Shiting Huang, Zecheng Li, Yu Zeng, Qingnan Ren, Zhen Fang, Qisheng Su, Kou Shi, Lin Chen, Zehui Chen, Feng Zhao
Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as an effective approach for enhancing the reasoning capabilities of Large Language Models (LLMs). Despite its efficacy, RLVR faces a meta-learning bottleneck: it lacks mechanisms for error attribution and experience internalization intrinsic to the human learning cycle beyond practice and verification, thereby limiting fine-grained credit assignment and reusable knowledge formation. We term such reusable knowledge representations derived from past errors as meta-experience. Based on this insight, we propose Meta-Experience Learning (MEL), a novel framework that incorporates self-distilled meta-experience into the model's parametric memory. Building upon standard RLVR, we introduce an additional design that leverages the LLM's self-verification capability to conduct contrastive analysis on paired correct and incorrect trajectories, identify the precise bifurcation points where reasoning errors arise, and summarize them into generalizable meta-experience. The meta-experience is further internalized into the LLM's parametric memory by minimizing the negative log-likelihood, which induces a language-modeled reward signal that bridges correct and incorrect reasoning trajectories and facilitates effective knowledge reuse. Experimental results demonstrate that MEL achieves consistent improvements on benchmarks, yielding 3.92%--4.73% Pass@1 gains across varying model sizes.
Authors: Ivana Nikoloska
Abstract: Quantum operations with indefinite causal order (ICO) represent a framework in quantum information processing where the relative order between two events can be indefinite. In this paper, we investigate whether sensing and computation, two canonical tasks in quantum information processing, can be carried out within the ICO framework. We propose a scheme for integrated sensing and computation that uses the same quantum state for both tasks. The quantum state is represented as an agent that performs state observation and learns a function of the state to make predictions via a parametric model. Under an ICO operation, the agent experiences a superposition of orders, one in which it performs state observation and then executes the required computation steps, and another in which the agent carries out the computation first and then performs state observation. This is distinct from prevailing information processing and machine intelligence paradigms where information acquisition and learning follow a strict causal order, with the former always preceding the latter. We provide experimental results and we show that the proposed scheme can achieve small training and testing losses on a representative task in magnetic navigation.
Authors: Haochen Wang, Yi Wu, Daryl Chang, Li Wei, Lukasz Heldt
Abstract: Optimizing large-scale machine learning systems, such as recommendation models for global video platforms, requires navigating a massive hyperparameter search space and, more critically, designing sophisticated optimizers, architectures, and reward functions to capture nuanced user behaviors. Achieving substantial improvements in these areas is a non-trivial task, traditionally relying on extensive manual iterations to test new hypotheses. We propose a self-evolving system that leverages Large Language Models (LLMs), specifically those from Google's Gemini family, to autonomously generate, train, and deploy high-performing, complex model changes within an end-to-end automated workflow. The self-evolving system is comprised of an Offline Agent (Inner Loop) that performs high-throughput hypothesis generation using proxy metrics, and an Online Agent (Outer Loop) that validates candidates against delayed north star business metrics in live production. Our agents act as specialized Machine Learning Engineers (MLEs): they exhibit deep reasoning capabilities, discovering novel improvements in optimization algorithms and model architecture, and formulating innovative reward functions that target long-term user engagement. The effectiveness of this approach is demonstrated through several successful production launches at YouTube, confirming that autonomous, LLM-driven evolution can surpass traditional engineering workflows in both development velocity and model performance.
Authors: Kirill Pavlenko, Alexander Golubev, Simon Karasik, Boris Yangel
Abstract: Group Relative Policy Optimization (GRPO) assigns a single scalar advantage to all tokens in a completion. For structured generations with explicit segments and objectives, this couples unrelated reward signals across segments, leading to objective interference and misattributed credit. We propose Blockwise Advantage Estimation, a family of GRPO-compatible methods that assigns each objective its own advantage and applies it only to the tokens in the corresponding text block, reducing reliance on hand-designed scalar rewards and scaling naturally to additional objectives. A key challenge is estimating advantages for later blocks whose rewards are conditioned on sampled prefixes; standard unbiased approaches require expensive nested rollouts from intermediate states. Concretely, we introduce an Outcome-Conditioned Baseline that approximates intermediate state values using only within-group statistics by stratifying samples according to a prefix-derived intermediate outcome. On math tasks with uncertainty estimation, our method mitigates reward interference, is competitive with a state-of-the-art reward-designed approach, and preserves test-time gains from confidence-weighted ensembling. More broadly, it provides a modular recipe for optimizing sequential objectives in structured generations without additional rollouts.
Authors: Alexey Kravatskiy, Valentin Khrulkov, Ivan Oseledets
Abstract: Recent advances in LLM-guided evolutionary computation, particularly AlphaEvolve, have demonstrated remarkable success in discovering novel mathematical constructions and solving challenging optimization problems. In this article, we present ImprovEvolve, a simple yet effective technique for enhancing LLM-based evolutionary approaches such as AlphaEvolve. Given an optimization problem, the standard approach is to evolve program code that, when executed, produces a solution close to the optimum. We propose an alternative program parameterization that maintains the ability to construct optimal solutions while reducing the cognitive load on the LLM. Specifically, we evolve a program (implementing, e.g., a Python class with a prescribed interface) that provides the following functionality: (1) propose a valid initial solution, (2) improve any given solution in terms of fitness, and (3) perturb a solution with a specified intensity. The optimum can then be approached by iteratively applying improve() and perturb() with a scheduled intensity. We evaluate ImprovEvolve on challenging problems from the AlphaEvolve paper: hexagon packing in a hexagon and the second autocorrelation inequality. For hexagon packing, the evolved program achieves new state-of-the-art results for 11, 12, 15, and 16 hexagons; a lightly human-edited variant further improves results for 14, 17, and 23 hexagons. For the second autocorrelation inequality, the human-edited program achieves a new state-of-the-art lower bound of 0.96258, improving upon AlphaEvolve's 0.96102.
Authors: Zhengbing He
Abstract: Stop-and-go waves, as a major form of freeway traffic congestion, cause severe and long-lasting adverse effects, including reduced traffic efficiency, increased driving risks, and higher vehicle emissions. Amongst the highway traffic management strategies, jam-absorption driving (JAD), in which a dedicated vehicle performs "slow-in" and "fast-out" maneuvers before being captured by a stop-and-go wave, has been proposed as a potential method for preventing the propagation of such waves. However, most existing JAD strategies remain impractical mainly due to the lack of discussion regarding implementation vehicles and operational conditions. Inspired by real-world observations of police-car swerving behavior, this paper first introduces a Single-Vehicle Two-Detector Jam-Absorption Driving (SVDD-JAD) problem, and then proposes a practical JAD strategy that transforms such behavior into a maneuver capable of suppressing the propagation of an isolated stop-and-go wave. Five key parameters that significantly affect the proposed strategy, namely, JAD speed, inflow traffic speed, wave width, wave speed, and in-wave speed, are identified and systematically analyzed. Using a SUMO-based simulation as an illustrative example, we further demonstrate how these parameters can be measured in practice with two stationary roadside traffic detectors. The results show that the proposed JAD strategy successfully suppresses the propagation of a stop-and-go wave, without triggering a secondary wave. This paper is expected to take a significant step toward making JAD practical, advancing it from a theoretical concept to a feasible and implementable strategy. To promote reproducibility in the transportation domain, we have also open-sourced all the code on our GitHub repository https://github.com/gotrafficgo.
Authors: Mayur Akewar, Sandeep Madireddy, Dongsheng Luo, Janki Bhimani
Abstract: Solid State Drives (SSDs) are critical to datacenters, consumer platforms, and mission-critical systems. Yet diagnosing their performance and reliability is difficult because data are fragmented and time-disjoint, and existing methods demand large datasets and expert input while offering only limited insights. Degradation arises not only from shifting workloads and evolving architectures but also from environmental factors such as temperature, humidity, and vibration. We present KORAL, a knowledge driven reasoning framework that integrates Large Language Models (LLMs) with a structured Knowledge Graph (KG) to generate insights into SSD operations. Unlike traditional approaches that require extensive expert input and large datasets, KORAL generates a Data KG from fragmented telemetry and integrates a Literature KG that already organizes knowledge from literature, reports, and traces. This turns unstructured sources into a queryable graph and telemetry into structured knowledge, and both the Graphs guide the LLM to deliver evidence-based, explainable analysis aligned with the domain vocabulary and constraints. Evaluation using real production traces shows that the KORAL delivers expert-level diagnosis and recommendations, supported by grounded explanations that improve reasoning transparency, guide operator decisions, reduce manual effort, and provide actionable insights to improve service quality. To our knowledge, this is the first end-to-end system that combines LLMs and KGs for full-spectrum SSD reasoning including Descriptive, Predictive, Prescriptive, and What-if analysis. We release the generated SSD-specific KG to advance reproducible research in knowledge-based storage system analysis. GitHub Repository: https://github.com/Damrl-lab/KORAL
Authors: Robert Ganian, Viktoriia Korchemna
Abstract: We investigate the parameterized complexity of Bayesian Network Structure Learning (BNSL), a classical problem that has received significant attention in empirical but also purely theoretical studies. We follow up on previous works that have analyzed the complexity of BNSL w.r.t. the so-called superstructure of the input. While known results imply that BNSL is unlikely to be fixed-parameter tractable even when parameterized by the size of a vertex cover in the superstructure, here we show that a different kind of parameterization - notably by the size of a feedback edge set - yields fixed-parameter tractability. We proceed by showing that this result can be strengthened to a localized version of the feedback edge set, and provide corresponding lower bounds that complement previous results to provide a complexity classification of BNSL w.r.t. virtually all well-studied graph parameters. We then analyze how the complexity of BNSL depends on the representation of the input. In particular, while the bulk of past theoretical work on the topic assumed the use of the so-called non-zero representation, here we prove that if an additive representation can be used instead then BNSL becomes fixed-parameter tractable even under significantly milder restrictions to the superstructure, notably when parameterized by the treewidth alone. Last but not least, we show how our results can be extended to the closely related problem of Polytree Learning.
Authors: Sarah Harkins Dayton, Layal Bou Hamdan, Ioannis D. Schizas, David L. Boothe, Vasileios Maroulas
Abstract: This chapter explores neural networks, topological data analysis, and topological deep learning techniques, alongside statistical Bayesian methods, for processing images, time series, and graphs to maximize the potential of artificial intelligence in the military domain. Throughout the chapter, we highlight practical applications spanning image, video, audio, and time-series recognition, fraud detection, and link prediction for graphical data, illustrating how topology-aware and uncertainty-aware models can enhance robustness, interpretability, and generalization.
Authors: Zehua Ma, Hanhui Li, Zhenyu Xie, Xiaonan Luo, Michael Kampffmeyer, Feng Gao, Xiaodan Liang
Abstract: Generating 3D content from a single image remains a fundamentally challenging and ill-posed problem due to the inherent absence of geometric and textural information in occluded regions. While state-of-the-art generative models can synthesize auxiliary views to provide additional supervision, these views inevitably contain geometric inconsistencies and textural misalignments that propagate and amplify artifacts during 3D reconstruction. To effectively harness these imperfect supervisory signals, we propose an adaptive optimization framework guided by excess risk decomposition, termed ERGO. Specifically, ERGO decomposes the optimization losses in 3D Gaussian splatting into two components, i.e., excess risk that quantifies the suboptimality gap between current and optimal parameters, and Bayes error that models the irreducible noise inherent in synthesized views. This decomposition enables ERGO to dynamically estimate the view-specific excess risk and adaptively adjust loss weights during optimization. Furthermore, we introduce geometry-aware and texture-aware objectives that complement the excess-risk-derived weighting mechanism, establishing a synergistic global-local optimization paradigm. Consequently, ERGO demonstrates robustness against supervision noise while consistently enhancing both geometric fidelity and textural quality of the reconstructed 3D content. Extensive experiments on the Google Scanned Objects dataset and the OmniObject3D dataset demonstrate the superiority of ERGO over existing state-of-the-art methods.
Authors: Jiqun Liu, Nischal Dinesh, Ran Yu
Abstract: ECHO (Evaluation of Chat, Human behavior, and Outcomes) is an open research platform designed to support reproducible, mixed-method studies of human interaction with both conversational AI systems and Web search engines. It enables researchers from varying disciplines to orchestrate end-to-end experimental workflows that integrate consent and background surveys, chat-based and search-based information-seeking sessions, writing or judgment tasks, and pre- and post-task evaluations within a unified, low-coding-load framework. ECHO logs fine-grained interaction traces and participant responses, and exports structured datasets for downstream analysis. By supporting both chat and search alongside flexible evaluation instruments, ECHO lowers technical barriers for studying learning, decision making, and user experience across different information access paradigms, empowering researchers from information retrieval, HCI, and the social sciences to conduct scalable and reproducible human-centered AI evaluations.
Authors: Mateo Juliani, Mingxuan Li, Elias Bareinboim
Abstract: Reward shaping has been applied widely to accelerate Reinforcement Learning (RL) agents' training. However, a principled way of designing effective reward shaping functions, especially for complex continuous control problems, remains largely under-explained. In this work, we propose to automatically learn a reward shaping function for continuous control problems from offline datasets, potentially contaminated by unobserved confounding variables. Specifically, our method builds upon the recently proposed causal Bellman equation to learn a tight upper bound on the optimal state values, which is then used as the potentials in the Potential-Based Reward Shaping (PBRS) framework. Our proposed reward shaping algorithm is tested with Soft-Actor-Critic (SAC) on multiple commonly used continuous control benchmarks and exhibits strong performance guarantees under unobserved confounders. More broadly, our work marks a solid first step towards confounding robust continuous control from a causal perspective. Code for training our reward shaping functions can be found at https://github.com/mateojuliani/confounding_robust_cont_control.
URLs: https://github.com/mateojuliani/confounding_robust_cont_control.
Authors: Zhimin Hu, Riya Roshan, Sashank Varma
Abstract: Human reasoning is shaped by resource rationality -- optimizing performance under constraints. Recently, inference-time scaling has emerged as a powerful paradigm to improve the reasoning performance of Large Language Models by expanding test-time computation. Specifically, instruction-tuned (IT) models explicitly generate long reasoning steps during inference, whereas Large Reasoning Models (LRMs) are trained by reinforcement learning to discover reasoning paths that maximize accuracy. However, it remains unclear whether resource-rationality can emerge from such scaling without explicit reward related to computational costs. We introduce a Variable Attribution Task in which models infer which variables determine outcomes given candidate variables, input-output trials, and predefined logical functions. By varying the number of candidate variables and trials, we systematically manipulate task complexity. Both models exhibit a transition from brute-force to analytic strategies as complexity increases. IT models degrade on XOR and XNOR functions, whereas LRMs remain robust. These findings suggest that models can adjust their reasoning behavior in response to task complexity, even without explicit cost-based reward. It provides compelling evidence that resource rationality is an emergent property of inference-time scaling itself.
Authors: Keenan Pepper, Alex McKenzie, Florin Pop, Stijn Servaes, Martin Leitgab, Mike Vaiana, Judd Rosenblatt, Michael S. A. Graziano, Diogo de Lucena
Abstract: Self-interpretation methods prompt language models to describe their own internal states, but remain unreliable due to hyperparameter sensitivity. We show that training lightweight adapters on interpretability artifacts, while keeping the LM entirely frozen, yields reliable self-interpretation across tasks and model families. A scalar affine adapter with just $d_\text{model}+1$ parameters suffices: trained adapters generate sparse autoencoder feature labels that outperform the training labels themselves (71% vs 63% generation scoring at 70B scale), identify topics with 94% recall@1 versus 1% for untrained baselines, and decode bridge entities in multi-hop reasoning that appear in neither prompt nor response, surfacing implicit reasoning without chain-of-thought. The learned bias vector alone accounts for 85% of improvement, and simpler adapters generalize better than more expressive alternatives. Controlling for model knowledge via prompted descriptions, we find self-interpretation gains outpace capability gains from 7B to 72B parameters. Our results demonstrate that self-interpretation improves with scale, without modifying the model being interpreted.
Authors: Jineel H Raythatha, Shuchang Ye, Jeremy Hsu, Jinman Kim
Abstract: Purpose: Translating foundation models into clinical practice requires evaluating their performance under compound distribution shift, where severe class imbalance coexists with heterogeneous imaging appearances. This challenge is relevant for traumatic bowel injury, a rare but high-mortality diagnosis. We investigated whether specificity deficits in foundation models are associated with heterogeneity in the negative class. Methods: This retrospective study used the multi-institutional, RSNA Abdominal Traumatic Injury CT dataset (2019-2023), comprising scans from 23 centres. Two foundation models (MedCLIP, zero-shot; RadDINO, linear probe) were compared against three task-specific approaches (CNN, Transformer, Ensemble). Models were trained on 3,147 patients (2.3% bowel injury prevalence) and evaluated on an enriched 100-patient test set. To isolate negative-class effects, specificity was assessed in patients without bowel injury who had concurrent solid organ injury (n=58) versus no abdominal pathology (n=50). Results: Foundation models achieved equivalent discrimination to task-specific models (AUC, 0.64-0.68 versus 0.58-0.64) with higher sensitivity (79-91% vs 41-74%) but lower specificity (33-50% vs 50-88%). All models demonstrated high specificity in patients without abdominal pathology (84-100%). When solid organ injuries were present, specificity declined substantially for foundation models (50-51 percentage points) compared with smaller reductions of 12-41 percentage points for task-specific models. Conclusion: Foundation models matched task-specific discrimination without task-specific training, but their specificity deficits were driven primarily by confounding negative-class heterogeneity rather than prevalence alone. Susceptibility to negative-class heterogeneity decreased progressively with labelled training, suggesting adaptation is required before clinical implementation.
Authors: Reese Kneeland, Wangshu Jiang, Ugo Bruzadin Nunes, Paul Steven Scotti, Arnaud Delorme, Jonathan Xu
Abstract: To be practical for real-life applications, models for brain-computer interfaces must be easily and quickly deployable on new subjects, effective on affordable scanning hardware, and small enough to run locally on accessible computing resources. To directly address these current limitations, we introduce ENIGMA, a multi-subject electroencephalography (EEG)-to-Image decoding model that reconstructs seen images from EEG recordings and achieves state-of-the-art (SOTA) performance on the research-grade THINGS-EEG2 and consumer-grade AllJoined-1.6M benchmarks, while fine-tuning effectively on new subjects with as little as 15 minutes of data. ENIGMA boasts a simpler architecture and requires less than 1% of the trainable parameters necessary for previous approaches. Our approach integrates a subject-unified spatio-temporal backbone along with a set of multi-subject latent alignment layers and an MLP projector to map raw EEG signals to a rich visual latent space. We evaluate our approach using a broad suite of image reconstruction metrics that have been standardized in the adjacent field of fMRI-to-Image research, and we describe the first EEG-to-Image study to conduct extensive behavioral evaluations of our reconstructions using human raters. Our simple and robust architecture provides a significant performance boost across both research-grade and consumer-grade EEG hardware, and a substantial improvement in fine-tuning efficiency and inference cost. Finally, we provide extensive ablations to determine the architectural choices most responsible for our performance gains in both single and multi-subject cases across multiple benchmark datasets. Collectively, our work provides a substantial step towards the development of practical brain-computer interface applications.
Authors: Mahmud Elahi Akhter, Federico Ruggeri, Iman Munire Bilal, Rob Procter, Maria Liakata
Abstract: Structured claim decomposition is often proposed as a solution for verifying complex, multi-faceted claims, yet empirical results have been inconsistent. We argue that these inconsistencies stem from two overlooked bottlenecks: evidence alignment and sub-claim error profiles. To better understand these factors, we introduce a new dataset of real-world complex claims, featuring temporally bounded evidence and human-annotated sub-claim evidence spans. We evaluate decomposition under two evidence alignment setups: Sub-claim Aligned Evidence (SAE) and Repeated Claim-level Evidence (SRE). Our results reveal that decomposition brings significant performance improvement only when evidence is granular and strictly aligned. By contrast, standard setups that rely on repeated claim-level evidence (SRE) fail to improve and often degrade performance as shown across different datasets and domains (PHEMEPlus, MMM-Fact, COVID-Fact). Furthermore, we demonstrate that in the presence of noisy sub-claim labels, the nature of the error ends up determining downstream robustness. We find that conservative "abstention" significantly reduces error propagation compared to aggressive but incorrect predictions. These findings suggest that future claim decomposition frameworks must prioritize precise evidence synthesis and calibrate the label bias of sub-claim verification models.
Authors: Jia Li, Yu Hou, Rui Zhang
Abstract: Automatically discovering personalized sequential events from large-scale time-series data is crucial for enabling precision medicine in clinical research, yet it remains a formidable challenge even for contemporary AI models. For example, while transformers capture rich associations, they are mostly agnostic to event timing and ordering, thereby bypassing potential causal reasoning. Intuitively, we need a method capable of evaluating the "degree of alignment" among patient-specific trajectories and identifying their shared patterns, i.e., the significant events in a consistent sequence. This necessitates treating timing as a true \emph{computable} dimension, allowing models to assign ``relative timestamps'' to candidate events beyond their observed physical times. In this work, we introduce LITT, a novel Timing-Transformer architecture that enables temporary alignment of sequential events on a virtual ``relative timeline'', thereby enabling \emph{event-timing-focused attention} and personalized interpretations of clinical trajectories. Its interpretability and effectiveness are validated on real-world longitudinal EHR data from 3,276 breast cancer patients to predict the onset timing of cardiotoxicity-induced heart disease. Furthermore, LITT outperforms both the benchmark and state-of-the-art survival analysis methods on public datasets, positioning it as a significant step forward for precision medicine in clinical AI.
Authors: Mehmet Hamza Erol, Xiangpeng Hao, Federico Bianchi, Ciro Greco, Jacopo Tagliabue, James Zou
Abstract: Traditional query optimization relies on cost-based optimizers that estimate execution cost (e.g., runtime, memory, and I/O) using predefined heuristics and statistical models. Improving these heuristics requires substantial engineering effort, and even when implemented, these heuristics often cannot take into account semantic correlations in queries and schemas that could enable better physical plans. Using our DBPlanBench harness for the DataFusion engine, we expose the physical plan through a compact serialized representation and let the LLM propose localized edits that can be applied and executed. We then apply an evolutionary search over these edits to refine candidates across iterations. Our key insight is that LLMs can leverage semantic knowledge to identify and apply non-obvious optimizations, such as join orderings that minimize intermediate cardinalities. We obtain up to 4.78$\times$ speedups on some queries and we demonstrate a small-to-large workflow in which optimizations found on small databases transfer effectively to larger databases.
Authors: Zhongzhi Li, Xuansheng Wu, Yijiang Li, Lijie Hu, Ninghao Liu
Abstract: The diversity of post-training data is critical for effective downstream performance in large language models (LLMs). Many existing approaches to constructing post-training data quantify diversity using text-based metrics that capture linguistic variation, but such metrics provide only weak signals for the task-relevant features that determine downstream performance. In this work, we introduce Feature Activation Coverage (FAC) which measures data diversity in an interpretable feature space. Building upon this metric, we further propose a diversity-driven data synthesis framework, named FAC Synthesis, that first uses a sparse autoencoder to identify missing features from a seed dataset, and then generates synthetic samples that explicitly reflect these features. Experiments show that our approach consistently improves both data diversity and downstream performance on various tasks, including instruction following, toxicity detection, reward modeling, and behavior steering. Interestingly, we identify a shared, interpretable feature space across model families (i.e., LLaMA, Mistral, and Qwen), enabling cross-model knowledge transfer. Our work provides a solid and practical methodology for exploring data-centric optimization of LLMs.
Authors: Khimya Khetarpal, Gheorghe Comanici, Jonathan Richens, Jeremy Shar, Fei Xia, Laurent Orseau, Aleksandra Faust, Doina Precup
Abstract: Full models of the world require complex knowledge of immense detail. While pre-trained large models have been hypothesized to contain similar knowledge due to extensive pre-training on vast amounts of internet scale data, using them directly in a search procedure is inefficient and inaccurate. Conversely, partial models focus on making high quality predictions for a subset of state and actions: those linked through affordances that achieve user intents~\citep{khetarpal2020can}. Can we posit large models as partial world models? We provide a formal answer to this question, proving that agents achieving task-agnostic, language-conditioned intents necessarily possess predictive partial-world models informed by affordances. In the multi-task setting, we introduce distribution-robust affordances and show that partial models can be extracted to significantly improve search efficiency. Empirical evaluations in tabletop robotics tasks demonstrate that our affordance-aware partial models reduce the search branching factor and achieve higher rewards compared to full world models.
Authors: Jiayun Pang, Ahmed M. Zaitoun, Xacobe Couso Cambeiro, Ivan Vuli\'c
Abstract: Adapting large language models (LLMs) trained on broad organic chemistry to smaller, domain-specific reaction datasets is a key challenge in chemical and pharmaceutical R&D. Effective specialisation requires learning new reaction knowledge while preserving general chemical understanding across related tasks. Here, we evaluate Low-Rank Adaptation (LoRA) as a parameter-efficient alternative to full fine-tuning for organic reaction prediction on limited, complex datasets. Using USPTO reaction classes and challenging C-H functionalisation reactions, we benchmark forward reaction prediction, retrosynthesis and reagent prediction. LoRA achieves accuracy comparable to full fine-tuning while effectively mitigating catastrophic forgetting and better preserving multi-task performance. Both fine-tuning approaches generalise beyond training distributions, producing plausible alternative solvent predictions. Notably, C-H functionalisation fine-tuning reveals that LoRA and full fine-tuning encode subtly different reactivity patterns, suggesting more effective reaction-specific adaptation with LoRA. As LLMs continue to scale, our results highlight the practicality of modular, parameter-efficient fine-tuning strategies for their flexible deployment for chemistry applications.
Authors: Alex Bie, Travis Dick, Alex Kulesza, Prabhakar Raghavan, Vinod Raman, Sergei Vassilvitskii
Abstract: Modern AI systems have been successfully deployed to win medals at international math competitions, assist with research workflows, and prove novel technical lemmas. However, despite their progress at advanced levels of mathematics, they remain stubbornly bad at basic arithmetic, consistently failing on the simple task of adding two numbers. We present a systematic investigation of this phenomenon. We demonstrate empirically that all frontier models suffer significantly degraded accuracy for integer addition as the number of digits increases. Furthermore, we show that most errors made by these models are highly interpretable and can be attributed to either operand misalignment or a failure to correctly carry; these two error classes explain 87.9%, 62.9%, and 92.4% of Claude Opus 4.1, GPT-5, and Gemini 2.5 Pro errors, respectively. Finally, we show that misalignment errors are frequently related to tokenization, and that carrying errors appear largely as independent random failures.
Authors: Zhongyao Wang, Taoyong Cui, Jiawen Zou, Shufei Zhang, Bo Yan, Wanli Ouyang, Weimin Tan, Mao Su
Abstract: Uncertainty quantification (UQ) is critical for assessing the reliability of machine learning interatomic potentials (MLIPs) in molecular dynamics (MD) simulations, identifying extrapolation regimes and enabling uncertainty-aware workflows such as active learning for training dataset construction. Existing UQ approaches for MLIPs are often limited by high computational cost or suboptimal performance. Evidential deep learning (EDL) provides a theoretically grounded single-model alternative that determines both aleatoric and epistemic uncertainty in a single forward pass. However, extending evidential formulations from scalar targets to vector-valued quantities such as atomic forces introduces substantial challenges, particularly in maintaining statistical self-consistency under rotational transformations. To address this, we propose \textit{Equivariant Evidential Deep Learning for Interatomic Potentials} ($\text{e}^2$IP), a backbone-agnostic framework that models atomic forces and their uncertainty jointly by representing uncertainty as a full $3\times3$ symmetric positive definite covariance tensor that transforms equivariantly under rotations. Experiments on diverse molecular benchmarks show that $\text{e}^2$IP provides a stronger accuracy-efficiency-reliability balance than the non-equivariant evidential baseline and the widely used ensemble method. It also achieves better data efficiency through the fully equivariant architecture while retaining single-model inference efficiency.
Authors: Wenkai Fan, Shurui Zhang, Xiaolong Wang, Haowei Yang, Tsz Wai Chan, Xingyan Chen, Junquan Bi, Zirui Zhou, Jia Liu, Kani Chen
Abstract: AIvilization v0 is a publicly deployed large-scale artificial society that couples a resource-constrained sandbox economy with a unified LLM-agent architecture, aiming to sustain long-horizon autonomy while remaining executable under rapidly changing environment. To mitigate the tension between goal stability and reactive correctness, we introduce (i) a hierarchical branch-thinking planner that decomposes life goals into parallel objective branches and uses simulation-guided validation plus tiered re-planning to ensure feasibility; (ii) an adaptive agent profile with dual-process memory that separates short-term execution traces from long-term semantic consolidation, enabling persistent yet evolving identity; and (iii) a human-in-the-loop steering interface that injects long-horizon objectives and short commands at appropriate abstraction levels, with effects propagated through memory rather than brittle prompt overrides. The environment integrates physiological survival costs, non-substitutable multi-tier production, an AMM-based price mechanism, and a gated education-occupation system. Using high-frequency transactions from the platforms mature phase, we find stable markets that reproduce key stylized facts (heavy-tailed returns and volatility clustering) and produce structured wealth stratification driven by education and access constraints. Ablations show simplified planners can match performance on narrow tasks, while the full architecture is more robust under multi-objective, long-horizon settings, supporting delayed investment and sustained exploration.
Authors: Jie Jiang, Yusen Huo, Xiangxin Zhan, Changping Wang, Jun Zhang
Abstract: Policy-based Reinforcement Learning (RL) has established itself as the dominant paradigm in generative recommendation for optimizing sequential user interactions. However, when applied to offline historical logs, these methods suffer a critical failure: the dominance of low-quality data induces severe model collapse. We first establish the Divergence Theory of Repulsive Optimization, revealing that negative gradient updates inherently trigger exponential intensity explosion during off-policy training. This theory elucidates the inherent dilemma of existing methods, exposing their inability to reconcile variance reduction and noise imitation. To break this curse, we argue that the solution lies in rigorously identifying the latent high-quality distribution entangled within the noisy behavior policy. Accordingly, we reformulate the objective as an Optimistic Distributionally Robust Optimization (DRO) problem. Guided by this formulation, we propose Distributionally Robust Policy Optimization (DRPO). We prove that hard filtering is the exact solution to this DRO objective, enabling DRPO to optimally recover high-quality behaviors while strictly discarding divergence-inducing noise. Extensive experiments demonstrate that DRPO achieves state-of-the-art performance on mixed-quality recommendation benchmarks.
Authors: Enzo Nicolas Spotorno, Antonio Augusto Medeiros Frohlich
Abstract: Runtime quantification of vehicle operational intensity is essential for predictive maintenance and condition monitoring in commercial and heavy-duty fleets. Traditional metrics like mileage fail to capture mechanical burden, while unsupervised deep learning models detect statistical anomalies, typically transient surface shocks, but often conflate statistical stability with mechanical rest. We identify this as a critical blind spot: high-load steady states, such as hill climbing with heavy payloads, appear statistically normal yet impose significant drivetrain fatigue. To resolve this, we propose a Dual-Stream Architecture that fuses unsupervised learning for surface anomaly detection with macroscopic physics proxies for cumulative load estimation. This approach leverages low-frequency sensor data to generate a multi-dimensional health vector, distinguishing between dynamic hazards and sustained mechanical effort. Validated on a RISC-V embedded platform, the architecture demonstrates low computational overhead, enabling comprehensive, edge-based health monitoring on resource-constrained ECUs without the latency or bandwidth costs of cloud-based monitoring.
Authors: Seonglae Cho, Zekun Wu, Adriano Koshiyama
Abstract: Sparse autoencoders (SAEs) decompose language model activations into interpretable features, but existing methods reveal only which features activate, not which change model outputs when amplified. We introduce Control Reinforcement Learning (CRL), which trains a policy to select SAE features for steering at each token, producing interpretable intervention logs: the learned policy identifies features that change model outputs when amplified. Adaptive Feature Masking encourages diverse feature discovery while preserving singlefeature interpretability. The framework yields new analysis capabilities: branch point tracking locates tokens where feature choice determines output correctness; critic trajectory analysis separates policy limitations from value estimation errors; layer-wise comparison reveals syntactic features in early layers and semantic features in later layers. On Gemma-2 2B across MMLU, BBQ, GSM8K, HarmBench, and XSTest, CRL achieves improvements while providing per-token intervention logs. These results establish learned feature steering as a mechanistic interpretability tool that complements static feature analysis with dynamic intervention probes
Authors: Liyang Chen, Hongkai Chen, Yujun Cai, Sifan Li, Qingwen Ye, Yiwei Wang
Abstract: Large Audio Language Models (LALMs) have demonstrated strong capabilities in audio understanding and reasoning. However, their performance on fine grained auditory perception remains unreliable, and existing approaches largely rely on data intensive training to internalize perceptual abilities. We propose AudioRouter, a reinforcement learning framework that enables LALMs to improve audio understanding by learning when and how to use external audio tools. Rather than tightly coupling tool usage with audio reasoning, AudioRouter formulates tool use as an explicit decision making problem and optimizes a lightweight routing policy while keeping the underlying reasoning model frozen. Experimental results show that AudioRouter achieves substantial improvements on standard audio understanding benchmarks while requiring up to 600x less training data to learn tool usage compared with conventional training paradigms. These findings suggest that learning effective tool usage offers a data efficient and scalable alternative to internalizing perceptual abilities in LALMs.
Authors: Feiyu Pan, Tianbin Zhang, Aoqian Zhang, Yu Sun, Zheng Wang, Lixing Chen, Li Pan, Jianhua Li
Abstract: Modern data lakes have emerged as foundational platforms for large-scale machine learning, enabling flexible storage of heterogeneous data and structured analytics through table-oriented abstractions. Despite their growing importance, standardized benchmarks for evaluating machine learning performance in data lake environments remain scarce. To address this gap, we present LakeMLB (Data Lake Machine Learning Benchmark), designed for the most common multi-source, multi-table scenarios in data lakes. LakeMLB focuses on two representative multi-table scenarios, Union and Join, and provides three real-world datasets for each scenario, covering government open data, finance, Wikipedia, and online marketplaces. The benchmark supports three representative integration strategies: pre-training-based, data augmentation-based, and feature augmentation-based approaches. We conduct extensive experiments with state-of-the-art tabular learning methods, offering insights into their performance under complex data lake scenarios. We release both datasets and code to facilitate rigorous research on machine learning in data lake ecosystems; the benchmark is available at https://github.com/zhengwang100/LakeMLB.
Authors: Pingbang Hu, Yuzheng Hu, Jiaqi W. Ma, Han Zhao
Abstract: Influence functions and related data attribution scores take the form of $g^{\top}F^{-1}g^{\prime}$, where $F\succeq 0$ is a curvature operator. In modern overparameterized models, forming or inverting $F\in\mathbb{R}^{d\times d}$ is prohibitive, motivating scalable influence computation via random projection with a sketch $P \in \mathbb{R}^{m\times d}$. This practice is commonly justified via the Johnson--Lindenstrauss (JL) lemma, which ensures approximate preservation of Euclidean geometry for a fixed dataset. However, JL does not address how sketching behaves under inversion. Furthermore, there is no existing theory that explains how sketching interacts with other widely-used techniques, such as ridge regularization and structured curvature approximations. We develop a unified theory characterizing when projection provably preserves influence functions. When $g,g^{\prime}\in\text{range}(F)$, we show that: 1) Unregularized projection: exact preservation holds iff $P$ is injective on $\text{range}(F)$, which necessitates $m\geq \text{rank}(F)$; 2) Regularized projection: ridge regularization fundamentally alters the sketching barrier, with approximation guarantees governed by the effective dimension of $F$ at the regularization scale; 3) Factorized influence: for Kronecker-factored curvatures $F=A\otimes E$, the guarantees continue to hold for decoupled sketches $P=P_A\otimes P_E$, even though such sketches exhibit row correlations that violate i.i.d. assumptions. Beyond this range-restricted setting, we analyze out-of-range test gradients and quantify a \emph{leakage} term that arises when test gradients have components in $\ker(F)$. This yields guarantees for influence queries on general test points. Overall, this work develops a novel theory that characterizes when projection provably preserves influence and provides principled guidance for choosing the sketch size in practice.
Authors: Zhong Li, Hongliang Lu, Tao Wei, Wenyu Liu, Yuxuan Chen, Yuan Lan, Fan Zhang, Zaiwen Wen
Abstract: Optimization modeling underpins decision-making in logistics, manufacturing, energy, and finance, yet translating natural-language requirements into correct optimization formulations and solver-executable code remains labor-intensive. Although large language models (LLMs) have been explored for this task, evaluation is still dominated by toy-sized or synthetic benchmarks, masking the difficulty of industrial problems with $10^{3}$--$10^{6}$ (or more) variables and constraints. A key bottleneck is the lack of benchmarks that align natural-language specifications with reference formulations/solver code grounded in real optimization models. To fill in this gap, we introduce MIPLIB-NL, built via a structure-aware reverse construction methodology from real mixed-integer linear programs in MIPLIB~2017. Our pipeline (i) recovers compact, reusable model structure from flat solver formulations, (ii) reverse-generates natural-language specifications explicitly tied to this recovered structure under a unified model--data separation format, and (iii) performs iterative semantic validation through expert review and human--LLM interaction with independent reconstruction checks. This yields 223 one-to-one reconstructions that preserve the mathematical content of the original instances while enabling realistic natural-language-to-optimization evaluation. Experiments show substantial performance degradation on MIPLIB-NL for systems that perform strongly on existing benchmarks, exposing failure modes invisible at toy scale.
Authors: Mohan Rajagopalan, Vinay Rao
Abstract: Agentic AI systems automate enterprise workflows but existing defenses--guardrails, semantic filters--are probabilistic and routinely bypassed. We introduce authenticated workflows, the first complete trust layer for enterprise agentic AI. Security reduces to protecting four fundamental boundaries: prompts, tools, data, and context. We enforce intent (operations satisfy organizational policies) and integrity (operations are cryptographically authentic) at every boundary crossing, combining cryptographic elimination of attack classes with runtime policy enforcement. This delivers deterministic security--operations either carry valid cryptographic proof or are rejected. We introduce MAPL, an AI-native policy language that expresses agentic constraints dynamically as agents evolve and invocation context changes, scaling as O(log M + N) policies versus O(M x N) rules through hierarchical composition with cryptographic attestations for workflow dependencies. We prove practicality through a universal security runtime integrating nine leading frameworks (MCP, A2A, OpenAI, Claude, LangChain, CrewAI, AutoGen, LlamaIndex, Haystack) through thin adapters requiring zero protocol modifications. Formal proofs establish completeness and soundness. Empirical validation shows 100% recall with zero false positives across 174 test cases, protection against 9 of 10 OWASP Top 10 risks, and complete mitigation of two high impact production CVEs.
Authors: Haoyu Hu, Raja Marjieh, Katherine M Collins, Chenyi Li, Thomas L. Griffiths, Ilia Sucholutsky, Nori Jacoby
Abstract: Writing code has been one of the most transformative ways for human societies to translate abstract ideas into tangible technologies. Modern AI is transforming this process by enabling experts and non-experts alike to generate code without actually writing code, but instead, through natural language instructions, or "vibe coding". While increasingly popular, the cumulative impact of vibe coding on productivity and collaboration, as well as the role of humans in this process, remains unclear. Here, we introduce a controlled experimental framework for studying collaborative vibe coding and use it to compare human-led, AI-led, and hybrid groups. Across 16 experiments involving 604 human participants, we show that people provide uniquely effective high-level instructions for vibe coding across iterations, whereas AI-provided instructions often result in performance collapse. We further demonstrate that hybrid systems perform best when humans retain directional control (providing the instructions), while evaluation is delegated to AI.
Authors: Yili Shen, Xiangliang Zhang
Abstract: Recent advances in reaction prediction have achieved near-saturated accuracy on standard benchmarks (e.g., USPTO), yet most state-of-the-art models formulate the task as a one-shot mapping from reactants to products, offering limited insight into the underlying reaction process. Procedural alternatives introduce stepwise generation but often rely on mechanism-specific supervision, discrete symbolic edits, and computationally expensive inference. In this work, we propose LatentRxnFlow, a new reaction prediction paradigm that models reactions as continuous latent trajectories anchored at the thermodynamic product state. Built on Conditional Flow Matching, our approach learns time-dependent latent dynamics directly from standard reactant-product pairs, without requiring mechanistic annotations or curated intermediate labels. While LatentRxnFlow achieves state-of-the-art performance on USPTO benchmarks, more importantly, the continuous formulation exposes the full generative trajectory, enabling trajectory-level diagnostics that are difficult to realize with discrete or one-shot models. We show that latent trajectory analysis allows us to localize and characterize failure modes and to mitigate certain errors via gated inference. Furthermore, geometric properties of the learned trajectories provide an intrinsic signal of epistemic uncertainty, helping prioritize reliably predictable reaction outcomes and flag ambiguous cases for additional validation. Overall, LatentRxnFlow combines strong predictive accuracy with improved transparency, diagnosability, and uncertainty awareness, moving reaction prediction toward more trustworthy deployment in high-throughput discovery workflows.
Authors: Mohan Rajagopalan, Vinay Rao
Abstract: Large Language Model (LLM) applications are vulnerable to prompt injection and context manipulation attacks that traditional security models cannot prevent. We introduce two novel primitives--authenticated prompts and authenticated context--that provide cryptographically verifiable provenance across LLM workflows. Authenticated prompts enable self-contained lineage verification, while authenticated context uses tamper-evident hash chains to ensure integrity of dynamic inputs. Building on these primitives, we formalize a policy algebra with four proven theorems providing protocol-level Byzantine resistance--even adversarial agents cannot violate organizational policies. Five complementary defenses--from lightweight resource controls to LLM-based semantic validation--deliver layered, preventative security with formal guarantees. Evaluation against representative attacks spanning 6 exhaustive categories achieves 100% detection with zero false positives and nominal overhead. We demonstrate the first approach combining cryptographically enforced prompt lineage, tamper-evident context, and provable policy reasoning--shifting LLM security from reactive detection to preventative guarantees.
Authors: Wei Chen, Xingyu Guo, Shuang Li, Zhao Zhang, Yan Zhong, Fuzhen Zhuang, Deqing wang
Abstract: Graph Domain Adaptation (GDA) transfers knowledge from labeled source graphs to unlabeled target graphs but is challenged by complex, multi-faceted distributional shifts. Existing methods attempt to reduce distributional shifts by aligning manually selected graph elements (e.g., node attributes or structural statistics), which typically require manually designed graph filters to extract relevant features before alignment. However, such approaches are inflexible: they rely on scenario-specific heuristics, and struggle when dominant discrepancies vary across transfer scenarios. To address these limitations, we propose \textbf{ADAlign}, an Adaptive Distribution Alignment framework for GDA. Unlike heuristic methods, ADAlign requires no manual specification of alignment criteria. It automatically identifies the most relevant discrepancies in each transfer and aligns them jointly, capturing the interplay between attributes, structures, and their dependencies. This makes ADAlign flexible, scenario-aware, and robust to diverse and dynamically evolving shifts. To enable this adaptivity, we introduce the Neural Spectral Discrepancy (NSD), a theoretically principled parametric distance that provides a unified view of cross-graph shifts. NSD leverages neural characteristic function in the spectral domain to encode feature-structure dependencies of all orders, while a learnable frequency sampler adaptively emphasizes the most informative spectral components for each task via minimax paradigm. Extensive experiments on 10 datasets and 16 transfer tasks show that ADAlign not only outperforms state-of-the-art baselines but also achieves efficiency gains with lower memory usage and faster training.
Authors: Yongzhong Xu
Abstract: We investigate the geometric structure of learning dynamics in overparameterized transformer models through carefully controlled modular arithmetic tasks. Our primary finding is that despite operating in high-dimensional parameter spaces ($d=128$), transformer training trajectories rapidly collapse onto low-dimensional execution manifolds of dimension $3$--$4$. This dimensional collapse is robust across random seeds and moderate task difficulties, though the orientation of the manifold in parameter space varies between runs. We demonstrate that this geometric structure underlies several empirically observed phenomena: (1) sharp attention concentration emerges as saturation along routing coordinates within the execution manifold, (2) stochastic gradient descent (SGD) exhibits approximately integrable dynamics when projected onto the execution subspace, with non-integrability confined to orthogonal staging directions, and (3) sparse autoencoders capture auxiliary routing structure but fail to isolate execution itself, which remains distributed across the low-dimensional manifold. Our results suggest a unifying geometric framework for understanding transformer learning, where the vast majority of parameters serve to absorb optimization interference while core computation occurs in a dramatically reduced subspace. These findings have implications for interpretability, training curriculum design, and understanding the role of overparameterization in neural network learning.
Authors: Wei Chen, Xingyu Guo, Shuang Li, Yan Zhong, Zhao Zhang, Fuzhen Zhuang, Hongrui Liu, Libang Zhang, Guo Ye, Huimei He
Abstract: Graph Domain Adaptation (GDA) aims to bridge distribution shifts between domains by transferring knowledge from well-labeled source graphs to given unlabeled target graphs. One promising recent approach addresses graph transfer by discretizing the adaptation process, typically through the construction of intermediate graphs or stepwise alignment procedures. However, such discrete strategies often fail in real-world scenarios, where graph structures evolve continuously and nonlinearly, making it difficult for fixed-step alignment to approximate the actual transformation process. To address these limitations, we propose \textbf{DiffGDA}, a \textbf{Diff}usion-based \textbf{GDA} method that models the domain adaptation process as a continuous-time generative process. We formulate the evolution from source to target graphs using stochastic differential equations (SDEs), enabling the joint modeling of structural and semantic transitions. To guide this evolution, a domain-aware network is introduced to steer the generative process toward the target domain, encouraging the diffusion trajectory to follow an optimal adaptation path. We theoretically show that the diffusion process converges to the optimal solution bridging the source and target domains in the latent space. Extensive experiments on 14 graph transfer tasks across 8 real-world datasets demonstrate DiffGDA consistently outperforms state-of-the-art baselines.
Authors: Dongshuo Yin, Xue Yang, Deng-Ping Fan, Shi-Min Hu
Abstract: Deploying vision foundation models typically relies on efficient adaptation strategies, whereas conventional full fine-tuning suffers from prohibitive costs and low efficiency. While delta-tuning has proven effective in boosting the performance and efficiency of LLMs during adaptation, its advantages cannot be directly transferred to the fine-tuning pipeline of vision foundation models. To push the boundaries of adaptation efficiency for vision tasks, we propose an adapter with Complex Linear Projection Optimization (CoLin). For architecture, we design a novel low-rank complex adapter that introduces only about 1% parameters to the backbone. For efficiency, we theoretically prove that low-rank composite matrices suffer from severe convergence issues during training, and address this challenge with a tailored loss. Extensive experiments on object detection, segmentation, image classification, and rotated object detection (remote sensing scenario) demonstrate that CoLin outperforms both full fine-tuning and classical delta-tuning approaches with merely 1% parameters for the first time, providing a novel and efficient solution for deployment of vision foundation models. We release the code on https://github.com/DongshuoYin/CoLin.
Authors: Shihao Dong, Yeke Chen, Zeren Luo, Jiahui Zhang, Bowen Xu, Jinghan Lin, Yimin Han, Ji Ma, Zhiyou Yu, Yudong Zhao, Peng Lu
Abstract: While single-agent legged locomotion has witnessed remarkable progress, individual robots remain fundamentally constrained by physical actuation limits. To transcend these boundaries, we introduce Co-jump, a cooperative task where two quadrupedal robots synchronize to execute jumps far beyond their solo capabilities. We tackle the high-impulse contact dynamics of this task under a decentralized setting, achieving synchronization without explicit communication or pre-specified motion primitives. Our framework leverages Multi-Agent Proximal Policy Optimization (MAPPO) enhanced by a progressive curriculum strategy, which effectively overcomes the sparse-reward exploration challenges inherent in mechanically coupled systems. We demonstrate robust performance in simulation and successful transfer to physical hardware, executing multi-directional jumps onto platforms up to 1.5 m in height. Specifically, one of the robots achieves a foot-end elevation of 1.1 m, which represents a 144% improvement over the 0.45 m jump height of a standalone quadrupedal robot, demonstrating superior vertical performance. Notably, this precise coordination is achieved solely through proprioceptive feedback, establishing a foundation for communication-free collaborative locomotion in constrained environments.
Authors: George Pu, Michael S. Lee, Udari Madhushani Sehwag, David J. Lee, Bryan Zhu, Yash Maurya, Mohit Raghavendra, Yuan Xue, Samuel Marc Denton
Abstract: Long-horizon workflow agents that operate effectively over extended periods are essential for truly autonomous systems. Their reliable execution critically depends on the ability to reason through ambiguous situations in which clarification seeking is necessary to ensure correct task execution. However, progress is limited by the lack of scalable, task-agnostic frameworks for systematically curating and measuring the impact of ambiguity across custom workflows. We address this gap by introducing LHAW (Long-Horizon Augmented Workflows), a modular, dataset-agnostic synthetic pipeline that transforms any well-specified task into controllable underspecified variants by systematically removing information across four dimensions - Goals, Constraints, Inputs, and Context - at configurable severity levels. Unlike approaches that rely on LLM predictions of ambiguity, LHAW validates variants through empirical agent trials, classifying them as outcome-critical, divergent, or benign based on observed terminal state divergence. We release 285 task variants from TheAgentCompany, SWE-Bench Pro and MCP-Atlas according to our taxonomy alongside formal analysis measuring how current agents detect, reason about, and resolve underspecification across ambiguous settings. LHAW provides the first systematic framework for cost-sensitive evaluation of agent clarification behavior in long-horizon settings, enabling development of reliable autonomous systems.
Authors: Scott P. McGrath, Katherine K. Kim, Karnjit Johl, Haibo Wang, Nick Anderson
Abstract: The integration of artificial intelligence (AI) into healthcare is accelerating, yet medical education has not kept pace with these technological advancements. This paper synthesizes current knowledge on AI in medical education through a comprehensive analysis of the literature, identifying key competencies, curricular approaches, and implementation strategies. The aim is highlighting the critical need for structured AI education across the medical learning continuum and offer a framework for curriculum development. The findings presented suggest that effective AI education requires longitudinal integration throughout medical training, interdisciplinary collaboration, and balanced attention to both technical fundamentals and clinical applications. This paper serves as a foundation for medical educators seeking to prepare future physicians for an AI-enhanced healthcare environment.
Authors: Seongwon Jin, Hanseul Choi, Sunggu Yang, Sungho Park, Jibum Kim
Abstract: Electroencephalography (ECoG) offers a promising alternative to conventional electrocorticography (EEG) for the early prediction of Parkinson's disease (PD), providing higher spatial resolution and a broader frequency range. However, reproducible comparisons has been limited by ethical constraints in human studies and the lack of open benchmark datasets. To address this gap, we introduce a new dataset, the first reproducible benchmark for PD prediction. It is constructed from long-term ECoG recordings of 6-hydroxydopamine (6-OHDA)-induced rat models and annotated with neural responses measured before and after electrical stimulation. In addition, we propose a Swap-Adversarial Framework (SAF) that mitigates high inter-subject variability and the high-dimensional low-sample-size (HDLSS) problem in ECoG data, while achieving robust domain generalization across ECoG and EEG-based Brain-Computer Interface (BCI) datasets. The framework integrates (1) robust preprocessing, (2) Inter-Subject Balanced Channel Swap (ISBCS) for cross-subject augmentation, and (3) domain-adversarial training to suppress subject-specific bias. ISBCS randomly swaps channels between subjects to reduce inter-subject variability, and domain-adversarial training jointly encourages the model to learn task-relevant shared features. We validated the effectiveness of the proposed method through extensive experiments under cross-subject, cross-session, and cross-dataset settings. Our method consistently outperformed all baselines across all settings, showing the most significant improvements in highly variable environments. Furthermore, the proposed method achieved superior cross-dataset performance between public EEG benchmarks, demonstrating strong generalization capability not only within ECoG but to EEG data. The new dataset and source code will be made publicly available upon publication.
Authors: Yuxin Ma, Nan Chen, Mateo D\'iaz, Soufiane Hayou, Dmitriy Kunisky, Soledad Villar
Abstract: Modern large-scale neural networks are often trained and released in multiple sizes to accommodate diverse inference budgets. To improve efficiency, recent work has explored model upscaling: initializing larger models from trained smaller ones in order to transfer knowledge and accelerate convergence. However, this method can be sensitive to hyperparameters that need to be tuned at the target upscaled model size, which is prohibitively costly to do directly. It remains unclear whether the most common workaround -- tuning on smaller models and extrapolating via hyperparameter scaling laws -- is still sound when using upscaling. We address this with principled approaches to upscaling with respect to model widths and efficiently tuning hyperparameters in this setting. First, motivated by $\mu$P and any-dimensional architectures, we introduce a general upscaling method applicable to a broad range of architectures and optimizers, backed by theory guaranteeing that models are equivalent to their widened versions and allowing for rigorous analysis of infinite-width limits. Second, we extend the theory of $\mu$Transfer to a hyperparameter transfer technique for models upscaled using our method and empirically demonstrate that this method is effective on realistic datasets and architectures.
Authors: Hanzhe Yu, Yun Ye, Jintao Rong, Qi Xuan, Chen Ma
Abstract: The rapid advancement of generative AI has raised concerns about the authenticity of digital images, as highly realistic fake images can now be generated at low cost, potentially increasing societal risks. In response, several datasets have been established to train detection models aimed at distinguishing AI-generated images from real ones. However, existing datasets suffer from limited generalization, low image quality, overly simple prompts, and insufficient image diversity. To address these limitations, we propose a high-quality, large-scale dataset comprising over 730,000 images across multiple categories, including both real and AI-generated images. The generated images are synthesized via state-of-the-art methods, including text-to-image generation (guided by over 10,000 carefully designed prompts), image inpainting, image refinement, and face swapping. Each generated image is annotated with its generation method and category. Inpainting images further include binary masks to indicate inpainted regions, providing rich metadata for analysis. Compared to existing datasets, detection models trained on our dataset demonstrate superior generalization capabilities. Our dataset not only serves as a strong benchmark for evaluating detection methods but also contributes to advancing the robustness of AI-generated image detection techniques. Building upon this, we propose a lightweight detection method based on image noise entropy, which transforms the original image into an entropy tensor of Non-Local Means (NLM) noise before classification. Extensive experiments demonstrate that models trained on our dataset achieve strong generalization, and our method delivers competitive performance, establishing a solid baseline for future research. The dataset and source code are publicly available at https://real-hd.github.io.
Authors: Shengyang Sun, Jiashen Hua, Junyi Feng, Xiaojin Gong
Abstract: Weakly supervised multimodal video anomaly detection has gained significant attention, yet the potential of the text modality remains under-explored. Text provides explicit semantic information that can enhance anomaly characterization and reduce false alarms. However, extracting effective text features is challenging due to the inability of general-purpose language models to capture anomaly-specific nuances and the scarcity of relevant descriptions. Furthermore, multimodal fusion often suffers from redundancy and imbalance. To address these issues, we propose a novel text-guided framework. First, we introduce an in-context learning-based multi-stage text augmentation mechanism to generate high-quality anomaly text samples for fine-tuning the text feature extractor. Second, we design a multi-scale bottleneck Transformer fusion module that uses compressed bottleneck tokens to progressively integrate information across modalities, mitigating redundancy and imbalance. Experiments on UCF-Crime and XD-Violence demonstrate state-of-the-art performance.
Authors: Guanting Ye, Qiyan Zhao, Wenhao Yu, Xiaofeng Zhang, Jianmin Ji, Yanyong Zhang, Ka-Veng Yuen
Abstract: Recent advances in 3D Large Multimodal Models (LMMs) built on Large Language Models (LLMs) have established the alignment of 3D visual features with LLM representations as the dominant paradigm. However, the inherited Rotary Position Embedding (RoPE) introduces limitations for multimodal processing. Specifically, applying 1D temporal positional indices disrupts the continuity of visual features along the column dimension, resulting in spatial locality loss. Moreover, RoPE follows the prior that temporally closer image tokens are more causally related, leading to long-term decay in attention allocation and causing the model to progressively neglect earlier visual tokens as the sequence length increases. To address these issues, we propose C^2RoPE, an improved RoPE that explicitly models local spatial Continuity and spatial Causal relationships for visual processing. C^2RoPE introduces a spatio-temporal continuous positional embedding mechanism for visual tokens. It first integrates 1D temporal positions with Cartesian-based spatial coordinates to construct a triplet hybrid positional index, and then employs a frequency allocation strategy to encode spatio-temporal positional information across the three index components. Additionally, we introduce Chebyshev Causal Masking, which determines causal dependencies by computing the Chebyshev distance of image tokens in 2D space. Evaluation results across various benchmarks, including 3D scene reasoning and 3D visual question answering, demonstrate C^2RoPE's effectiveness. The code is be available at https://github.com/ErikZ719/C2RoPE.
Authors: Junichiro Takahashi, Masataka Sato, Satoshi Kodeta, Norihiko Takeda
Abstract: Recent advances in large language models (LLMs) have enabled the development of multimodal medical AI. While models such as MedGemini achieve high accuracy on VQA tasks like USMLE MM, their performance on ECG based tasks remains limited, and some models, such as MedGemma, do not support ECG data at all. Interpreting ECGs is inherently challenging, and diagnostic accuracy can vary depending on the interpreter's experience. Although echocardiography provides rich diagnostic information, it requires specialized equipment and personnel, limiting its availability. In this study, we focus on constructing a robust ECG encoder for multimodal pretraining using real world hospital data. We employ SigLIP, a CLIP based model with a sigmoid based loss function enabling multi label prediction, and introduce a modified loss function tailored to the multi label nature of ECG data. Experiments demonstrate that incorporating medical knowledge in the language model and applying the modified loss significantly improve multi label ECG classification. To further enhance performance, we increase the embedding dimensionality and apply random cropping to mitigate data drift. Finally, per label analysis reveals which ECG findings are easier or harder to predict. Our study provides a foundational framework for developing medical models that utilize ECG data.
Authors: Lihan Zha, Asher J. Hancock, Mingtong Zhang, Tenny Yin, Yixuan Huang, Dhruv Shah, Allen Z. Ren, Anirudha Majumdar
Abstract: A long-standing goal in robotics is a generalist policy that can be deployed zero-shot on new robot embodiments without per-embodiment adaptation. Despite large-scale multi-embodiment pre-training, existing Vision-Language-Action models (VLAs) remain tightly coupled to their training embodiments and typically require costly fine-tuning. We introduce Language-Action Pre-training (LAP), a simple recipe that represents low-level robot actions directly in natural language, aligning action supervision with the pre-trained vision-language model's input-output distribution. LAP requires no learned tokenizer, no costly annotation, and no embodiment-specific architectural design. Based on LAP, we present LAP-3B, which to the best of our knowledge is the first VLA to achieve substantial zero-shot transfer to previously unseen robot embodiments without any embodiment-specific fine-tuning. Across multiple novel robots and manipulation tasks, LAP-3B attains over 50% average zero-shot success, delivering roughly a 2x improvement over the strongest prior VLAs. We further show that LAP enables efficient adaptation and favorable scaling, while unifying action prediction and VQA in a shared language-action format that yields additional gains through co-training.
Authors: Leheng Sheng, Yongtao Zhang, Wenchang Ma, Yaorui Shi, Ting Huang, Xiang Wang, An Zhang, Ke Shen, Tat-Seng Chua
Abstract: While reasoning over long context is crucial for various real-world applications, it remains challenging for large language models (LLMs) as they suffer from performance degradation as the context length grows. Recent work MemAgent has tried to tackle this by processing context chunk-by-chunk in an RNN-like loop and updating a textual memory for final answering. However, this naive recurrent memory update faces two crucial drawbacks: (i) memory can quickly explode because it can update indiscriminately, even on evidence-free chunks; and (ii) the loop lacks an exit mechanism, leading to unnecessary computation after even sufficient evidence is collected. To address these issues, we propose GRU-Mem, which incorporates two text-controlled gates for more stable and efficient long-context reasoning. Specifically, in GRU-Mem, the memory only updates when the update gate is open and the recurrent loop will exit immediately once the exit gate is open. To endow the model with such capabilities, we introduce two reward signals $r^{\text{update}}$ and $r^{\text{exit}}$ within end-to-end RL, rewarding the correct updating and exiting behaviors respectively. Experiments on various long-context reasoning tasks demonstrate the effectiveness and efficiency of GRU-Mem, which generally outperforms the vanilla MemAgent with up to 400\% times inference speed acceleration.
Authors: Chenhao Zhang, Yazhe Niu, Hongsheng Li
Abstract: Metaphorical comprehension in images remains a critical challenge for Nowadays AI systems. While Multimodal Large Language Models (MLLMs) excel at basic Visual Question Answering (VQA), they consistently struggle to grasp the nuanced cultural, emotional, and contextual implications embedded in visual content. This difficulty stems from the task's demand for sophisticated multi-hop reasoning, cultural context, and Theory of Mind (ToM) capabilities, which current models lack. To fill this gap, we propose MetaphorStar, the first end-to-end visual reinforcement learning (RL) framework for image implication tasks. Our framework includes three core components: the fine-grained dataset TFQ-Data, the visual RL method TFQ-GRPO, and the well-structured benchmark TFQ-Bench. Our fully open-source MetaphorStar family, trained using TFQ-GRPO on TFQ-Data, significantly improves performance by an average of 82.6% on the image implication benchmarks. Compared with 20+ mainstream MLLMs, MetaphorStar-32B achieves state-of-the-art (SOTA) on Multiple-Choice Question and Open-Style Question, significantly outperforms the top closed-source model Gemini-3.0-pro on True-False Question. Crucially, our experiments reveal that learning image implication tasks improves the general understanding ability, especially the complex visual reasoning ability. We further provide a systematic analysis of model parameter scaling, training data scaling, and the impact of different model architectures and training strategies, demonstrating the broad applicability of our method. We open-sourced all model weights, datasets, and method code at https://metaphorstar.github.io.
Authors: Boxiao Wang, Kai Li, Tianyi Liu, Chen Li, Junzhe Wang, Yifan Zhang, Jian Cheng
Abstract: Symbolic regression aims to distill mathematical equations from observational data. Recent approaches have successfully leveraged Large Language Models (LLMs) to generate equation hypotheses, capitalizing on their vast pre-trained scientific priors. However, existing frameworks predominantly treat the LLM as a static generator, relying on prompt-level guidance to steer exploration. This paradigm fails to update the model's internal representations based on search feedback, often yielding physically inconsistent or mathematically redundant expressions. In this work, we propose PiT-PO (Physics-informed Token-regularized Policy Optimization), a unified framework that evolves the LLM into an adaptive generator via reinforcement learning. Central to PiT-PO is a dual-constraint mechanism that rigorously enforces hierarchical physical validity while simultaneously applying fine-grained, token-level penalties to suppress redundant structures. Consequently, PiT-PO aligns LLM to produce equations that are both scientifically consistent and structurally parsimonious. Empirically, PiT-PO achieves state-of-the-art performance on standard benchmarks and successfully discovers novel turbulence models for challenging fluid dynamics problems. We also demonstrate that PiT-PO empowers small-scale models to outperform closed-source giants, democratizing access to high-performance scientific discovery.
Authors: Guangzhi Xiong, Sanchit Sinha, Aidong Zhang
Abstract: The trade-off between interpretability and accuracy remains a core challenge in machine learning. Standard Generalized Additive Models (GAMs) offer clear feature attributions but are often constrained by their strictly additive nature, which can limit predictive performance. Introducing feature interactions can boost accuracy yet may obscure individual feature contributions. To address these issues, we propose Neural Additive Experts (NAEs), a novel framework that seamlessly balances interpretability and accuracy. NAEs employ a mixture of experts framework, learning multiple specialized networks per feature, while a dynamic gating mechanism integrates information across features, thereby relaxing rigid additive constraints. Furthermore, we propose targeted regularization techniques to mitigate variance among expert predictions, facilitating a smooth transition from an exclusively additive model to one that captures intricate feature interactions while maintaining clarity in feature attributions. Our theoretical analysis and experiments on synthetic data illustrate the model's flexibility, and extensive evaluations on real-world datasets confirm that NAEs achieve an optimal balance between predictive accuracy and transparent, feature-level explanations. The code is available at https://github.com/Teddy-XiongGZ/NAE.
Authors: Ailin Huang, Ang Li, Aobo Kong, Bin Wang, Binxing Jiao, Bo Dong, Bojun Wang, Boyu Chen, Brian Li, Buyun Ma, Chang Su, Changxin Miao, Changyi Wan, Chao Lou, Chen Hu, Chen Xu, Chenfeng Yu, Chengting Feng, Chengyuan Yao, Chunrui Han, Dan Ma, Dapeng Shi, Daxin Jiang, Dehua Ma, Deshan Sun, Di Qi, Enle Liu, Fajie Zhang, Fanqi Wan, Guanzhe Huang, Gulin Yan, Guoliang Cao, Guopeng Li, Han Cheng, Hangyu Guo, Hanshan Zhang, Hao Nie, Haonan Jia, Haoran Lv, Hebin Zhou, Hekun Lv, Heng Wang, Heung-Yeung Shum, Hongbo Huang, Hongbo Peng, Hongyu Zhou, Hongyuan Wang, Houyong Chen, Huangxi Zhu, Huimin Wu, Huiyong Guo, Jia Wang, Jian Zhou, Jianjian Sun, Jiaoren Wu, Jiaran Zhang, Jiashu Lv, Jiashuo Liu, Jiayi Fu, Jiayu Liu, Jie Cheng, Jie Luo, Jie Yang, Jie Zhou, Jieyi Hou, Jing Bai, Jingcheng Hu, Jingjing Xie, Jingwei Wu, Jingyang Zhang, Jishi Zhou, Junfeng Liu, Junzhe Lin, Ka Man Lo, Kai Liang, Kaibo Liu, Kaijun Tan, Kaiwen Yan, Kaixiang Li, Kang An, Kangheng Lin, Lei Yang, Liang Lv, Liang Zhao, Liangyu Chen, Lieyu Shi, Liguo Tan, Lin Lin, Lina Chen, Luck Ma, Mengqiang Ren, Michael Li, Ming Li, Mingliang Li, Mingming Zhang, Mingrui Chen, Mitt Huang, Na Wang, Peng Liu, Qi Han, Qian Zhao, Qinglin He, Qinxin Du, Qiuping Wu, Quan Sun, Rongqiu Yang, Ruihang Miao, Ruixin Han, Ruosi Wan, Ruyan Guo, Shan Wang, Shaoliang Pang, Shaowen Yang, Shengjie Fan, Shijie Shang, Shiliang Yang, Shiwei Li, Shuangshuang Tian, Siqi Liu, Siye Wu, Siyu Chen, Song Yuan, Tiancheng Cao, Tianchi Yue, Tianhao Cheng, Tianning Li, Tingdan Luo, Wang You, Wei Ji, Wei Yuan, Wei Zhang, Weibo Wu, Weihao Xie, Wen Sun, Wenjin Deng, Wenzhen Zheng, Wuxun Xie, Xiangfeng Wang, Xiangwen Kong, Xiangyu Liu, Xiangyu Zhang, Xiaobo Yang, Xiaojia Liu, Xiaolan Yuan, Xiaoran Jiao, Xiaoxiao Ren, Xiaoyun Zhang, Xin Li, Xin Liu, Xin Wu, Xing Chen, Xingping Yang, Xinran Wang, Xu Zhao, Xuan He, Xuanti Feng, Xuedan Cai, Xuqiang Zhou, Yanbo Yu, Yang Li, Yang Xu, Yanlin Lai, Yanming Xu, Yaoyu Wang, Yeqing Shen, Yibo Zhu, Yichen Lv, Yicheng Cao, Yifeng Gong, Yijing Yang, Yikun Yang, Yin Zhao, Yingxiu Zhao, Yinmin Zhang, Yitong Zhang, Yixuan Zhang, Yiyang Chen, Yongchi Zhao, Yongshen Long, Yongyao Wang, Yousong Guan, Yu Zhou, Yuang Peng, Yuanhao Ding, Yuantao Fan, Yuanzhen Yang, Yuchu Luo, Yudi Zhao, Yue Peng, Yueqiang Lin, Yufan Lu, Yuling Zhao, Yunzhou Ju, Yurong Zhang, Yusheng Li, Yuxiang Yang, Yuyang Chen, Yuzhu Cai, Zejia Weng, Zetao Hong, Zexi Li, Zhe Xie, Zheng Ge, Zheng Gong, Zheng Zeng, Zhenyi Lu, Zhewei Huang, Zhichao Chang, Zhiguo Huang, Zhiheng Hu, Zidong Yang, Zili Wang, Ziqi Ren, Zixin Zhang, Zixuan Wang
Abstract: We introduce Step 3.5 Flash, a sparse Mixture-of-Experts (MoE) model that bridges frontier-level agentic intelligence and computational efficiency. We focus on what matters most when building agents: sharp reasoning and fast, reliable execution. Step 3.5 Flash pairs a 196B-parameter foundation with 11B active parameters for efficient inference. It is optimized with interleaved 3:1 sliding-window/full attention and Multi-Token Prediction (MTP-3) to reduce the latency and cost of multi-round agentic interactions. To reach frontier-level intelligence, we design a scalable reinforcement learning framework that combines verifiable signals with preference feedback, while remaining stable under large-scale off-policy training, enabling consistent self-improvement across mathematics, code, and tool use. Step 3.5 Flash demonstrates strong performance across agent, coding, and math tasks, achieving 85.4% on IMO-AnswerBench, 86.4% on LiveCodeBench-v6 (2024.08-2025.05), 88.2% on tau2-Bench, 69.0% on BrowseComp (with context management), and 51.0% on Terminal-Bench 2.0, comparable to frontier models such as GPT-5.2 xHigh and Gemini 3.0 Pro. By redefining the efficiency frontier, Step 3.5 Flash provides a high-density foundation for deploying sophisticated agents in real-world industrial environments.
Authors: Sansheng Cao, Zhengyu Ma, Yonghong Tian
Abstract: Zeroth-order (ZO) optimization has long been favored for its biological plausibility and its capacity to handle non-differentiable objectives, yet its computational complexity has historically limited its application in deep neural networks. Challenging the conventional paradigm that gradients propagate layer-by-layer, we propose Hierarchical Zeroth-Order (HZO) optimization, a novel divide-and-conquer strategy that decomposes the depth dimension of the network. We prove that HZO reduces the query complexity from $O(ML^2)$ to $O(ML \log L)$ for a network of width $M$ and depth $L$, representing a significant leap over existing ZO methodologies. Furthermore, we provide a detailed error analysis showing that HZO maintains numerical stability by operating near the unitary limit ($L_{lip} \approx 1$). Extensive evaluations on CIFAR-10 and ImageNet demonstrate that HZO achieves competitive accuracy compared to backpropagation.
Authors: Shuo He, Lang Feng, Xin Cheng, Lei Feng, Bo An
Abstract: Reinforcement learning for large language models suffers from high-variance token-level importance sampling (IS) ratios, which would destabilize policy optimization at scale. To improve stability, recent methods typically use a fixed sequence-level IS ratio for all tokens in a sequence or adjust each token's IS ratio separately, thereby neglecting temporal off-policy derivation across tokens in a sequence. In this paper, we first empirically identify that local off-policy deviation is structurally inconsistent at the token level, which may distort policy-gradient updates across adjacent tokens and lead to training collapse. To address the issue, we propose Online Causal Kalman Filtering for stable and effective Policy Optimization (KPO). Concretely, we model the desired IS ratio as a latent state that evolves across tokens and apply a Kalman filter to update this state online and autoregressively based on the states of past tokens, regardless of future tokens. The resulting filtered IS ratios preserve token-wise local structure-aware variation while strongly smoothing noise spikes, yielding more stable and effective policy updates. Experimentally, KPO achieves superior results on challenging math reasoning datasets compared with state-of-the-art counterparts.
Authors: Zhibin Duan, Guowei Rong, Zhuo Li, Bo Chen, Mingyuan Zhou, Dandan Guo
Abstract: Reward models learned from human preferences are central to aligning large language models (LLMs) via reinforcement learning from human feedback, yet they are often vulnerable to reward hacking due to noisy annotations and systematic biases such as response length or style. We propose Bayesian Non-Negative Reward Model (BNRM), a principled reward modeling framework that integrates non-negative factor analysis into Bradley-Terry (BT) preference model. BNRM represents rewards through a sparse, non-negative latent factor generative process that operates at two complementary levels: instance-specific latent variables induce disentangled reward representations, while sparsity over global latent factors acts as an implicit debiasing mechanism that suppresses spurious correlations. Together, this disentanglement-then-debiasing structure enables robust uncertainty-aware reward learning. To scale BNRM to modern LLMs, we develop an amortized variational inference network conditioned on deep model representations, allowing efficient end-to-end training. Extensive empirical results demonstrate that BNRM substantially mitigates reward over-optimization, improves robustness under distribution shifts, and yields more interpretable reward decompositions than strong baselines.
Authors: Siyuan Yan, Xieji Li, Dan Mo, Philipp Tschandl, Yiwen Jiang, Zhonghua Wang, Ming Hu, Lie Ju, Cristina Vico-Alonso, Yizhen Zheng, Jiahe Liu, Juexiao Zhou, Camilla Chello, Jen G. Cheung, Julien Anriot, Luc Thomas, Clare Primiero, Gin Tan, Aik Beng Ng, Simon See, Xiaoying Tang, Albert Ip, Xiaoyang Liao, Adrian Bowling, Martin Haskett, Shuang Zhao, Monika Janda, H. Peter Soyer, Victoria Mar, Harald Kittler, Zongyuan Ge
Abstract: Medical foundation models have shown promise in controlled benchmarks, yet widespread deployment remains hindered by reliance on task-specific fine-tuning. Here, we introduce DermFM-Zero, a dermatology vision-language foundation model trained via masked latent modelling and contrastive learning on over 4 million multimodal data points. We evaluated DermFM-Zero across 20 benchmarks spanning zero-shot diagnosis and multimodal retrieval, achieving state-of-the-art performance without task-specific adaptation. We further evaluated its zero-shot capabilities in three multinational reader studies involving over 1,100 clinicians. In primary care settings, AI assistance enabled general practitioners to nearly double their differential diagnostic accuracy across 98 skin conditions. In specialist settings, the model significantly outperformed board-certified dermatologists in multimodal skin cancer assessment. In collaborative workflows, AI assistance enabled non-experts to surpass unassisted experts while improving management appropriateness. Finally, we show that DermFM-Zero's latent representations are interpretable: sparse autoencoders unsupervisedly disentangle clinically meaningful concepts that outperform predefined-vocabulary approaches and enable targeted suppression of artifact-induced biases, enhancing robustness without retraining. These findings demonstrate that a foundation model can provide effective, safe, and transparent zero-shot clinical decision support.
Authors: Suyash Mishra
Abstract: This white paper presents a critical synthesis of the recent breakthrough in nonuniformly elliptic regularity theory and the burgeoning field of neurosymbolic large reasoning models (LRMs). We explore the resolution of the long-standing sharp growth rate conjecture in Schauder theory, achieved by Cristiana De Filippis and Giuseppe Mingione, which identifies the exact threshold $q/p < 1 + \alpha/n$ for gradient H\"{o}lder continuity. Central to this mathematical achievement is the ``ghost equation'' methodology, a sophisticated auxiliary derivation that bypasses the non-differentiability of classical Euler-Lagrange systems. We propose that the next era of mathematical discovery lies in the integration of these pure analytical constructs with LRMs grounded in topos theory and formal verification frameworks such as Safe and Typed Chain-of-Thought (PC-CoT). By modeling the reasoning process as a categorical colimit in a slice topos, we demonstrate how LRMs can autonomously navigate the ``Dark Side'' of the calculus of variations, providing machine-checkable proofs for regularity bounds in complex, multi-phase physical systems.
Authors: Junhua Liu, Zhangcheng Wang, Zhike Han, Ningli Wang, Guotao Liang, Kun Kuang
Abstract: Visual Chain-of-Thought (VCoT) has emerged as a promising paradigm for enhancing multimodal reasoning by integrating visual perception into intermediate reasoning steps. However, existing VCoT approaches are largely confined to static scenarios and struggle to capture the temporal dynamics essential for tasks such as instruction, prediction, and camera motion. To bridge this gap, we propose TwiFF-2.7M, the first large-scale, temporally grounded VCoT dataset derived from $2.7$ million video clips, explicitly designed for dynamic visual question and answer. Accompanying this, we introduce TwiFF-Bench, a high-quality evaluation benchmark of $1,078$ samples that assesses both the plausibility of reasoning trajectories and the correctness of final answers in open-ended dynamic settings. Building on these foundations, we propose the TwiFF model, a unified modal that synergistically leverages pre-trained video generation and image comprehension capabilities to produce temporally coherent visual reasoning cues-iteratively generating future action frames and textual reasoning. Extensive experiments demonstrate that TwiFF significantly outperforms existing VCoT methods and Textual Chain-of-Thought baselines on dynamic reasoning tasks, which fully validates the effectiveness for visual question answering in dynamic scenarios. Our code and data is available at https://github.com/LiuJunhua02/TwiFF.
Authors: Jinjie Shen, Jing Wu, Yaxiong Wang, Lechao Cheng, Shengeng Tang, Tianrui Hui, Nan Pu, Zhun Zhong
Abstract: Existing forgery detection methods are often limited to uni-modal or bi-modal settings, failing to handle the interleaved text, images, and videos prevalent in real-world misinformation. To bridge this gap, this paper targets to develop a unified framework for omnibus vision-language forgery detection and grounding. In this unified setting, the {interplay} between diverse modalities and the dual requirements of simultaneous detection and localization pose a critical ``difficulty bias`` problem: the simpler veracity classification task tends to dominate the gradients, leading to suboptimal performance in fine-grained grounding during multi-task optimization. To address this challenge, we propose \textbf{OmniVL-Guard}, a balanced reinforcement learning framework for omnibus vision-language forgery detection and grounding. Particularly, OmniVL-Guard comprises two core designs: Self-Evolving CoT Generatio and Adaptive Reward Scaling Policy Optimization (ARSPO). {Self-Evolving CoT Generation} synthesizes high-quality reasoning paths, effectively overcoming the cold-start challenge. Building upon this, {Adaptive Reward Scaling Policy Optimization (ARSPO)} dynamically modulates reward scales and task weights, ensuring a balanced joint optimization. Extensive experiments demonstrate that OmniVL-Guard significantly outperforms state-of-the-art methods and exhibits zero-shot robust generalization across out-of-domain scenarios.
Authors: Guobin Shen, Chenxiao Zhao, Xiang Cheng, Lei Huang, Xing Yu
Abstract: Training stability remains a central challenge in reinforcement learning (RL) for large language models (LLMs). Policy staleness, asynchronous training, and mismatches between training and inference engines all cause the behavior policy to diverge from the current policy, risking training collapse. Importance sampling provides a principled correction for this distribution shift but suffers from high variance; existing remedies such as token-level clipping and sequence-level normalization lack a unified theoretical foundation. We propose Variational sEquence-level Soft Policy Optimization (VESPO). By incorporating variance reduction into a variational formulation over proposal distributions, VESPO derives a closed-form reshaping kernel that operates directly on sequence-level importance weights without length normalization. Experiments on mathematical reasoning benchmarks show that VESPO maintains stable training under staleness ratios up to 64x and fully asynchronous execution, and delivers consistent gains across both dense and Mixture-of-Experts models. Code is available at https://github.com/FloyedShen/VESPO
Authors: Zhifeng Rao, Wenlong Chen, Lei Xie, Xia Hua, Dongfu Yin, Zhen Tian, F. Richard Yu
Abstract: Vision-Language-Action (VLA) models have recently achieved remarkable progress in robotic perception and control, yet most existing approaches primarily rely on VLM trained using 2D images, which limits their spatial understanding and action grounding in complex 3D environments. To address this limitation, we propose a novel framework that integrates depth estimation into VLA models to enrich 3D feature representations. Specifically, we employ a depth estimation baseline called VGGT to extract geometry-aware 3D cues from standard RGB inputs, enabling efficient utilization of existing large-scale 2D datasets while implicitly recovering 3D structural information. To further enhance the reliability of these depth-derived features, we introduce a new module called action assistant, which constrains the learned 3D representations with action priors and ensures their consistency with downstream control tasks. By fusing the enhanced 3D features with conventional 2D visual tokens, our approach significantly improves the generalization ability and robustness of VLA models. Experimental results demonstrate that the proposed method not only strengthens perception in geometrically ambiguous scenarios but also leads to superior action prediction accuracy. This work highlights the potential of depth-driven data augmentation and auxiliary expert supervision for bridging the gap between 2D observations and 3D-aware decision-making in robotic systems.
Authors: Qiuran Zhao, Kai Ming Ting, Xinpeng Li
Abstract: The task of graph-level anomaly detection (GLAD) is to identify anomalous graphs that deviate significantly from the majority of graphs in a dataset. While deep GLAD methods have shown promising performance, their black-box nature limits their reliability and deployment in real-world applications. Although some recent methods have made attempts to provide explanations for anomaly detection results, they either provide explanations without referencing normal graphs, or rely on abstract latent vectors as prototypes rather than concrete graphs from the dataset. To address these limitations, we propose Prototype-based Graph-Level Anomaly Detection (ProtoGLAD), an interpretable unsupervised framework that provides explanation for each detected anomaly by explicitly contrasting with its nearest normal prototype graph. It employs a point-set kernel to iteratively discover multiple normal prototype graphs and their associated clusters from the dataset, then identifying graphs distant from all discovered normal clusters as anomalies. Extensive experiments on multiple real-world datasets demonstrate that ProtoGLAD achieves competitive anomaly detection performance compared to state-of-the-art GLAD methods while providing better human-interpretable prototype-based explanations.
Authors: Hyeongmin Lee, Chanyeol Choi, Jihoon Kwon, Yoon Kim, Alejandro Lopez-Lira, Wonbin Ahn, Yongjae Lee
Abstract: Asset retrieval--finding similar assets in a financial universe--is central to quantitative investment decision-making. Existing approaches define similarity through historical price patterns or sector classifications, but such backward-looking criteria provide no guarantee about future behavior. We argue that effective asset retrieval should be future-aligned: the retrieved assets should be those most likely to exhibit correlated future returns. To this end, we propose Future-Aligned Soft Contrastive Learning (FASCL), a representation learning framework whose soft contrastive loss uses pairwise future return correlations as continuous supervision targets. We further introduce an evaluation protocol designed to directly assess whether retrieved assets share similar future trajectories. Experiments on 4,229 US equities demonstrate that FASCL consistently outperforms 13 baselines across all future-behavior metrics. The source code will be available soon.
Authors: Yifei Li, Weidong Guo, Lingling Zhang, Rongman Xu, Muye Huang, Hui Liu, Lijiao Xu, Yu Xu, Jun Liu
Abstract: Long-term conversational memory is a core capability for LLM-based dialogue systems, yet existing benchmarks and evaluation protocols primarily focus on surface-level factual recall. In realistic interactions, appropriate responses often depend on implicit constraints such as user state, goals, or values that are not explicitly queried later. To evaluate this setting, we introduce \textbf{LoCoMo-Plus}, a benchmark for assessing cognitive memory under cue--trigger semantic disconnect, where models must retain and apply latent constraints across long conversational contexts. We further show that conventional string-matching metrics and explicit task-type prompting are misaligned with such scenarios, and propose a unified evaluation framework based on constraint consistency. Experiments across diverse backbone models, retrieval-based methods, and memory systems demonstrate that cognitive memory remains challenging and reveals failures not captured by existing benchmarks. Our code and evaluation framework are publicly available at: https://github.com/xjtuleeyf/Locomo-Plus.
Authors: Davide Evangelista, Pasquale Cascarano, Elena Loli Piccolomini
Abstract: The reconstruction of X-rays CT images from sparse or limited-angle geometries is a highly challenging task. The lack of data typically results in artifacts in the reconstructed image and may even lead to object distortions. For this reason, the use of deep generative models in this context has great interest and potential success. In the Deep Generative Prior (DGP) framework, the use of diffusion-based generative models is combined with an iterative optimization algorithm for the reconstruction of CT images from sinograms acquired under sparse geometries, to maintain the explainability of a model-based approach while introducing the generative power of a neural network. There are therefore several aspects that can be further investigated within these frameworks to improve reconstruction quality, such as image generation, the model, and the iterative algorithm used to solve the minimization problem, for which we propose modifications with respect to existing approaches. The results obtained even under highly sparse geometries are very promising, although further research is clearly needed in this direction.
Authors: Hugo L. Hammer, Vajira Thambawita, P{\aa}l Halvorsen
Abstract: A narrated e-book combines synchronized audio with digital text, highlighting the currently spoken word or sentence during playback. This format supports early literacy and assists individuals with reading challenges, while also allowing general readers to seamlessly switch between reading and listening. With the emergence of natural-sounding neural Text-to-Speech (TTS) technology, several commercial services have been developed to leverage these technology for converting standard text e-books into high-quality narrated e-books. However, no open-source solutions currently exist to perform this task. In this paper, we present Calliope, an open-source framework designed to fill this gap. Our method leverages state-of-the-art open-source TTS to convert a text e-book into a narrated e-book in the EPUB 3 Media Overlay format. The method offers several innovative steps: audio timestamps are captured directly during TTS, ensuring exact synchronization between narration and text highlighting; the publisher's original typography, styling, and embedded media are strictly preserved; and the entire pipeline operates offline. This offline capability eliminates recurring API costs, mitigates privacy concerns, and avoids copyright compliance issues associated with cloud-based services. The framework currently supports the state-of-the-art open-source TTS systems XTTS-v2 and Chatterbox. A potential alternative approach involves first generating narration via TTS and subsequently synchronizing it with the text using forced alignment. However, while our method ensures exact synchronization, our experiments show that forced alignment introduces drift between the audio and text highlighting significant enough to degrade the reading experience. Source code and usage instructions are available at https://github.com/hugohammer/TTS-Narrated-Ebook-Creator.git.
URLs: https://github.com/hugohammer/TTS-Narrated-Ebook-Creator.git.
Authors: Kian Majlessi, Amir Masoud Soltani, Mohammad Ebrahim Mahdavi, Aurelien Gourrier, Peyman Adibi
Abstract: Super-resolution (SR) applied to real-world low-resolution (LR) images often results in complex, irregular degradations that stem from the inherent complexity of natural scene acquisition. In contrast to SR artifacts arising from synthetic LR images created under well-defined scenarios, those distortions are highly unpredictable and vary significantly across different real-life contexts. Consequently, assessing the quality of SR images (SR-IQA) obtained from realistic LR, remains a challenging and underexplored problem. In this work, we introduce a no-reference SR-IQA approach tailored for such highly ill-posed realistic settings. The proposed method enables domain-adaptive IQA for real-world SR applications, particularly in data-scarce domains. We hypothesize that degradations in super-resolved images are strongly dependent on the underlying SR algorithms, rather than being solely determined by image content. To this end, we introduce a self-supervised learning (SSL) strategy that first pretrains multiple SR model oriented representations in a pretext stage. Our contrastive learning framework forms positive pairs from images produced by the same SR model and negative pairs from those generated by different methods, independent of image content. The proposed approach S3 RIQA, further incorporates targeted preprocessing to extract complementary quality information and an auxiliary task to better handle the various degradation profiles associated with different SR scaling factors. To this end, we constructed a new dataset, SRMORSS, to support unsupervised pretext training; it includes a wide range of SR algorithms applied to numerous real LR images, which addresses a gap in existing datasets. Experiments on real SR-IQA benchmarks demonstrate that S3 RIQA consistently outperforms most state-of-the-art relevant metrics.
Authors: Rumman Firdos, Aman Dangi
Abstract: The growing sophistication of modern malware and phishing campaigns has diminished the effectiveness of traditional signature-based intrusion detection systems. This work presents SecureScan, an AI-driven, triple-layer detection framework that integrates logistic regression-based classification, heuristic analysis, and external threat intelligence via the VirusTotal API for comprehensive triage of URLs, file hashes, and binaries. The proposed architecture prioritizes efficiency by filtering known threats through heuristics, classifying uncertain samples using machine learning, and validating borderline cases with third-party intelligence. On benchmark datasets, SecureScan achieves 93.1 percent accuracy with balanced precision (0.87) and recall (0.92), demonstrating strong generalization and reduced overfitting through threshold-based decision calibration. A calibrated threshold and gray-zone logic (0.45-0.55) were introduced to minimize false positives and enhance real-world stability. Experimental results indicate that a lightweight statistical model, when augmented with calibrated verification and external intelligence, can achieve reliability and performance comparable to more complex deep learning systems.
Authors: Charlotte Cambier van Nooten, Christos Aronis, Yuliya Shapovalova, Lucia Cavallaro
Abstract: This paper explores sparsification methods as a form of regularization in Graph Neural Networks (GNNs) to address high memory usage and computational costs in large-scale graph applications. Using techniques from Network Science and Machine Learning, including Erd\H{o}s-R\'enyi for model sparsification, we enhance the efficiency of GNNs for real-world applications. We demonstrate our approach on N-1 contingency assessment in electrical grids, a critical task for ensuring grid reliability. We apply our methods to three datasets of varying sizes, exploring Graph Convolutional Networks (GCN) and Graph Isomorphism Networks (GIN) with different degrees of sparsification and rewiring. Comparison across sparsification levels shows the potential of combining insights from both research fields to improve GNN performance and scalability. Our experiments highlight the importance of tuning sparsity parameters: while sparsity can improve generalization, excessive sparsity may hinder learning of complex patterns. Our adaptive rewiring approach, particularly when combined with early stopping, proves promising by allowing the model to adapt its connectivity structure during training. This research contributes to understanding how sparsity can be effectively leveraged in GNNs for critical applications like power grid reliability analysis.
Authors: Bram Van Bolderik (Technische Universiteit Eindhoven, The Netherlands), Vlado Menkovski (Technische Universiteit Eindhoven, The Netherlands), Sonia Heemstra de Groot (Eindhoven Technical University, The Netherlands), Manil Dev Gomony (Eindhoven University of Technology, The Netherlands)
Abstract: Neural network based receivers have recently demonstrated superior system-level performance compared to traditional receivers. However, their practicality is limited by high memory and power requirements, as separate weight sets must be stored for each code rate. To address this challenge, we propose LOREN, a Low Rank-Based Code-Rate Adaptation Neural Receiver that achieves adaptability with minimal overhead. LOREN integrates lightweight low rank adaptation adapters (LOREN adapters) into convolutional layers, freezing a shared base network while training only small adapters per code rate. An end-to-end training framework over 3GPP CDL channels ensures robustness across realistic wireless environments. LOREN achieves comparable or superior performance relative to fully retrained base neural receivers. The hardware implementation of LOREN in 22nm technology shows more than 65% savings in silicon area and up to 15% power reduction when supporting three code rates.
Authors: Enrico Ahlers, Daniel Passon, Yannic Noller, Lars Grunske
Abstract: Machine learning models are increasingly present in our everyday lives; as a result, they become targets of adversarial attackers seeking to manipulate the systems we interact with. A well-known vulnerability is a backdoor introduced into a neural network by poisoned training data or a malicious training process. Backdoors can be used to induce unwanted behavior by including a certain trigger in the input. Existing mitigations filter training data, modify the model, or perform expensive input modifications on samples. If a vulnerable model has already been deployed, however, those strategies are either ineffective or inefficient. To address this gap, we propose our inference-time backdoor mitigation approach called FIRE (Feature-space Inference-time REpair). We hypothesize that a trigger induces structured and repeatable changes in the model's internal representation. We view the trigger as directions in the latent spaces between layers that can be applied in reverse to correct the inference mechanism. Therefore, we turn the backdoored model against itself by manipulating its latent representations and moving a poisoned sample's features along the backdoor directions to neutralize the trigger. Our evaluation shows that FIRE has low computational overhead and outperforms current runtime mitigations on image benchmarks across various attacks, datasets, and network architectures.
Authors: Samal Mukhtar, Yinghua Yao, Zhu Sun, Mustafa Mustafa, Yew Soon Ong, Youcheng Sun
Abstract: Software vulnerability detection (SVD) is a critical challenge in modern systems. Large language models (LLMs) offer natural-language explanations alongside predictions, but most work focuses on binary evaluation, and explanations often lack semantic consistency with Common Weakness Enumeration (CWE) categories. We propose VulReaD, a knowledge-graph-guided approach for vulnerability reasoning and detection that moves beyond binary classification toward CWE-level reasoning. VulReaD leverages a security knowledge graph (KG) as a semantic backbone and uses a strong teacher LLM to generate CWE-consistent contrastive reasoning supervision, enabling student model training without manual annotations. Students are fine-tuned with Odds Ratio Preference Optimization (ORPO) to encourage taxonomy-aligned reasoning while suppressing unsupported explanations. Across three real-world datasets, VulReaD improves binary F1 by 8-10% and multi-class classification by 30% Macro-F1 and 18% Micro-F1 compared to state-of-the-art baselines. Results show that LLMs outperform deep learning baselines in binary detection and that KG-guided reasoning enhances CWE coverage and interpretability.
Authors: Benjy Friedmann, Nadav Dym
Abstract: Recent advances in Neural Combinatorial Optimization (NCO) have been dominated by diffusion models that treat the Euclidean Traveling Salesman Problem (TSP) as a stochastic $N \times N$ heatmap generation task. In this paper, we propose CycFlow, a framework that replaces iterative edge denoising with deterministic point transport. CycFlow learns an instance-conditioned vector field that continuously transports input 2D coordinates to a canonical circular arrangement, where the optimal tour is recovered from this $2N$ dimensional representation via angular sorting. By leveraging data-dependent flow matching, we bypass the quadratic bottleneck of edge scoring in favor of linear coordinate dynamics. This paradigm shift accelerates solving speed by up to three orders of magnitude compared to state-of-the-art diffusion baselines, while maintaining competitive optimality gaps.
Authors: Zihui Zhou, Yong Feng, Yanying Chen, Guofan Duan, Zhenxi Song, Mingliang Zhou, Weijia Jia
Abstract: Multimodal large language models (MLLMs) are increasingly adopted in remote sensing (RS) and have shown strong performance on tasks such as RS visual grounding (RSVG), RS visual question answering (RSVQA), and multimodal dialogue. However, hallucinations, which are responses inconsistent with the input RS images, severely hinder their deployment in high-stakes scenarios (e.g., emergency management and agricultural monitoring) and remain under-explored in RS. In this work, we present RSHallu, a systematic study with three deliverables: (1) we formalize RS hallucinations with an RS-oriented taxonomy and introduce image-level hallucination to capture RS-specific inconsistencies beyond object-centric errors (e.g., modality, resolution, and scene-level semantics); (2) we build a hallucination benchmark RSHalluEval (2,023 QA pairs) and enable dual-mode checking, supporting high-precision cloud auditing and low-cost reproducible local checking via a compact checker fine-tuned on RSHalluCheck dataset (15,396 QA pairs); and (3) we introduce a domain-tailored dataset RSHalluShield (30k QA pairs) for training-friendly mitigation and further propose training-free plug-and-play strategies, including decoding-time logit correction and RS-aware prompting. Across representative RS-MLLMs, our mitigation improves the hallucination-free rate by up to 21.63 percentage points under a unified protocol, while maintaining competitive performance on downstream RS tasks (RSVQA/RSVG). Code and datasets will be released.
Authors: Rasmus Krebs, Somnath Mazumdar
Abstract: Recent studies have revealed that when LLMs are appropriately prompted and configured, they demonstrate mixed results. Such results often meet or exceed the baseline performance. However, these comparisons have two primary issues. First, they mostly considered only reliability as a comparison metric and selected a few LLMs (such as Codex and ChatGPT) for comparision. This paper proposes a comprehensive code quality assessment framework called Programmatic Excellence via LLM Iteration (PELLI). PELLI is an iterative analysis-based process that upholds high-quality code changes. We extended the state-of-the-art by performing a comprehensive evaluation that generates quantitative metrics for analyzing three primary nonfunctional requirements (such as maintainability, performance, and reliability) while selecting five popular LLMs. For PELLI's applicability, we selected three application domains while following Python coding standards. Following this framework, practitioners can ensure harmonious integration between LLMs and human developers, ensuring that their potential is fully realized. PELLI can serve as a practical guide for developers aiming to leverage LLMs while adhering to recognized quality standards. This study's outcomes are crucial for advancing LLM technologies in real-world applications, providing stakeholders with a clear understanding of where these LLMs excel and where they require further refinement. Overall, based on three nonfunctional requirements, we have found that GPT-4T and Gemini performed slightly better. We also found that prompt design can influence the overall code quality. In addition, each application domain demonstrated high and low scores across various metrics, and even within the same metrics across different prompts.
Authors: Deyuan Liu, Zecheng Wang, Zhanyue Qin, Zhiying Tu, Dianhui Chu, Dianbo Sui
Abstract: Large Language Models (LLMs) exhibit impressive capabilities yet suffer from sensitivity to slight input context variations, hampering reliability. Conventional metrics like accuracy and perplexity fail to assess local prediction robustness, as normalized output probabilities can obscure the underlying resilience of an LLM's internal state to perturbations. We introduce the Token Constraint Bound ($\delta_{\mathrm{TCB}}$), a novel metric that quantifies the maximum internal state perturbation an LLM can withstand before its dominant next-token prediction significantly changes. Intrinsically linked to output embedding space geometry, $\delta_{\mathrm{TCB}}$ provides insights into the stability of the model's internal predictive commitment. Our experiments show $\delta_{\mathrm{TCB}}$ correlates with effective prompt engineering and uncovers critical prediction instabilities missed by perplexity during in-context learning and text generation. $\delta_{\mathrm{TCB}}$ offers a principled, complementary approach to analyze and potentially improve the contextual stability of LLM predictions.
Authors: Yuexiao Ma, Xuzhe Zheng, Jing Xu, Xiwei Xu, Feng Ling, Xiawu Zheng, Huafeng Kuang, Huixia Li, Xing Wang, Xuefeng Xiao, Fei Chao, Rongrong Ji
Abstract: Autoregressive models, often built on Transformer architectures, represent a powerful paradigm for generating ultra-long videos by synthesizing content in sequential chunks. However, this sequential generation process is notoriously slow. While caching strategies have proven effective for accelerating traditional video diffusion models, existing methods assume uniform denoising across all frames-an assumption that breaks down in autoregressive models where different video chunks exhibit varying similarity patterns at identical timesteps. In this paper, we present FlowCache, the first caching framework specifically designed for autoregressive video generation. Our key insight is that each video chunk should maintain independent caching policies, allowing fine-grained control over which chunks require recomputation at each timestep. We introduce a chunkwise caching strategy that dynamically adapts to the unique denoising characteristics of each chunk, complemented by a joint importance-redundancy optimized KV cache compression mechanism that maintains fixed memory bounds while preserving generation quality. Our method achieves remarkable speedups of 2.38 times on MAGI-1 and 6.7 times on SkyReels-V2, with negligible quality degradation (VBench: 0.87 increase and 0.79 decrease respectively). These results demonstrate that FlowCache successfully unlocks the potential of autoregressive models for real-time, ultra-long video generation-establishing a new benchmark for efficient video synthesis at scale. The code is available at https://github.com/mikeallen39/FlowCache.
Authors: Fanpu Cao, Lu Dai, Jindong Han, Hui Xiong
Abstract: Multivariate time series forecasting (MTSF) plays a vital role in numerous real-world applications, yet existing models remain constrained by their reliance on a limited historical context. This limitation prevents them from effectively capturing global periodic patterns that often span cycles significantly longer than the input horizon - despite such patterns carrying strong predictive signals. Naive solutions, such as extending the historical window, lead to severe drawbacks, including overfitting, prohibitive computational costs, and redundant information processing. To address these challenges, we introduce the Global Temporal Retriever (GTR), a lightweight and plug-and-play module designed to extend any forecasting model's temporal awareness beyond the immediate historical context. GTR maintains an adaptive global temporal embedding of the entire cycle and dynamically retrieves and aligns relevant global segments with the input sequence. By jointly modeling local and global dependencies through a 2D convolution and residual fusion, GTR effectively bridges short-term observations with long-term periodicity without altering the host model architecture. Extensive experiments on six real-world datasets demonstrate that GTR consistently delivers state-of-the-art performance across both short-term and long-term forecasting scenarios, while incurring minimal parameter and computational overhead. These results highlight GTR as an efficient and general solution for enhancing global periodicity modeling in MTSF tasks. Code is available at this repository: https://github.com/macovaseas/GTR.
Authors: Luigi Simeone
Abstract: Time Series Foundation Models (TSFMs) have introduced zero-shot prediction capabilities that bypass the need for task-specific training. Whether these capabilities translate to mission-critical applications such as electricity demand forecasting--where accuracy, calibration, and robustness directly affect grid operations--remains an open question. We present a multi-dimensional benchmark evaluating four TSFMs (Chronos-Bolt, Chronos-2, Moirai-2, and TinyTimeMixer) alongside Prophet as an industry-standard baseline and two statistical references (SARIMA and Seasonal Naive), using ERCOT hourly load data from 2020 to 2024. All experiments run on consumer-grade hardware (AMD Ryzen 7, 16GB RAM, no GPU). The evaluation spans four axes: (1) context length sensitivity from 24 to 2048 hours, (2) probabilistic forecast calibration, (3) robustness under distribution shifts including COVID-19 lockdowns and Winter Storm Uri, and (4) prescriptive analytics for operational decision support. The top-performing foundation models achieve MASE values near 0.31 at long context lengths (C = 2048h, day-ahead horizon), a 47% reduction over the Seasonal Naive baseline. The inclusion of Prophet exposes a structural advantage of pre-trained models: Prophet fails when the fitting window is shorter than its seasonality period (MASE > 74 at 24-hour context), while TSFMs maintain stable accuracy even with minimal context because they recognise temporal patterns learned during pre-training rather than estimating them from scratch. Calibration varies substantially across models--Chronos-2 produces well-calibrated prediction intervals (95% empirical coverage at 90% nominal level) while both Moirai-2 and Prophet exhibit overconfidence (~70% coverage). We provide practical model selection guidelines and release the complete benchmark framework for reproducibility.
Authors: Cong Pang, Xuyu Feng, Yujie Yi, Zixuan Chen, Jiawei Hong, Tiankuo Yao, Nang Yuan, Jiapeng Luo, Lewei Lu, Xin Lou
Abstract: Despite the strong performance achieved by reinforcement learning-trained information-seeking agents, learning in open-ended web environments remains severely constrained by low signal-to-noise feedback. Text-based parsers often discard layout semantics and introduce unstructured noise, while long-horizon training typically relies on sparse outcome rewards that obscure which retrieval actions actually matter. We propose a visual-native search framework that represents webpages as visual snapshots, allowing agents to leverage layout cues to quickly localize salient evidence and suppress distractors. To learn effectively from these high-dimensional observations, we introduce Information-Aware Credit Assignment (ICA), a post-hoc method that estimates each retrieved snapshot's contribution to the final outcome via posterior analysis and propagates dense learning signals back to key search turns. Integrated with a GRPO-based training pipeline, our approach consistently outperforms text-based baselines on diverse information-seeking benchmarks, providing evidence that visual snapshot grounding with information-level credit assignment alleviates the credit-assignment bottleneck in open-ended web environments. The code and datasets will be released in https://github.com/pc-inno/ICA_MM_deepsearch.git.
Authors: Xuefeng Xu, Graham Cormode
Abstract: Federated Learning (FL) enables multiple parties to collaboratively train machine learning models without sharing raw data. However, before training, data must be preprocessed to address missing values, inconsistent formats, and heterogeneous feature scales. This preprocessing stage is critical for model performance but is largely overlooked in FL research. In practical FL systems, privacy constraints prohibit centralizing raw data, while communication efficiency introduces further challenges for distributed preprocessing. We introduce FedPS, a unified framework for federated data preprocessing based on aggregated statistics. FedPS leverages data-sketching techniques to efficiently summarize local datasets while preserving essential statistical information. Building on these summaries, we design federated algorithms for feature scaling, encoding, discretization, and missing-value imputation, and extend preprocessing-related models such as k-Means, k-Nearest Neighbors, and Bayesian Linear Regression to both horizontal and vertical FL settings. FedPS provides flexible, communication-efficient, and consistent preprocessing pipelines for practical FL deployments.
Authors: Zhiyin Tan, Jennifer D'Souza
Abstract: Systematic reviews and meta-analyses rely on converting narrative articles into structured, numerically grounded study records. Despite rapid advances in large language models (LLMs), it remains unclear whether they can meet the structural requirements of this process, which hinge on preserving roles, methods, and effect-size attribution across documents rather than on recognizing isolated entities. We propose a structural, diagnostic framework that evaluates LLM-based evidence extraction as a progression of schema-constrained queries with increasing relational and numerical complexity, enabling precise identification of failure points beyond atom-level extraction. Using a manually curated corpus spanning five scientific domains, together with a unified query suite and evaluation protocol, we evaluate two state-of-the-art LLMs under both per-document and long-context, multi-document input regimes. Across domains and models, performance remains moderate for single-property queries but degrades sharply once tasks require stable binding between variables, roles, statistical methods, and effect sizes. Full meta-analytic association tuples are extracted with near-zero reliability, and long-context inputs further exacerbate these failures. Downstream aggregation amplifies even minor upstream errors, rendering corpus-level statistics unreliable. Our analysis shows that these limitations stem not from entity recognition errors, but from systematic structural breakdowns, including role reversals, cross-analysis binding drift, instance compression in dense result sections, and numeric misattribution, indicating that current LLMs lack the structural fidelity, relational binding, and numerical grounding required for automated meta-analysis. The code and data are publicly available at GitHub (https://github.com/zhiyintan/LLM-Meta-Analysis).
Authors: Zhuohan Xie, Rania Elbadry, Fan Zhang, Georgi Georgiev, Xueqing Peng, Lingfei Qian, Jimin Huang, Dimitar Dimitrov, Vanshikaa Jani, Yuyang Dai, Jiahui Geng, Yuxia Wang, Ivan Koychev, Veselin Stoyanov, Preslav Nakov
Abstract: We present the setup and the tasks of the FinMMEval Lab at CLEF 2026, which introduces the first multilingual and multimodal evaluation framework for financial Large Language Models (LLMs). While recent advances in financial natural language processing have enabled automated analysis of market reports, regulatory documents, and investor communications, existing benchmarks remain largely monolingual, text-only, and limited to narrow subtasks. FinMMEval 2026 addresses this gap by offering three interconnected tasks that span financial understanding, reasoning, and decision-making: Financial Exam Question Answering, Multilingual Financial Question Answering (PolyFiQA), and Financial Decision Making. Together, these tasks provide a comprehensive evaluation suite that measures models' ability to reason, generalize, and act across diverse languages and modalities. The lab aims to promote the development of robust, transparent, and globally inclusive financial AI systems, with datasets and evaluation resources publicly released to support reproducible research.
Authors: Berfin Sakallioglu, Giorgia Nadizar, Eric Medvet
Abstract: Multi-task policy search is a challenging problem because policies are required to generalize beyond training cases. Curriculum learning has proven to be effective in this setting, as it introduces complexity progressively. However, designing effective curricula is labor-intensive and requires extensive domain expertise. LLM-based curriculum generation has only recently emerged as a potential solution, but was limited to operate in static, offline modes without leveraging real-time feedback from the optimizer. Here we propose an interactive LLM-assisted framework for online curriculum generation, where the LLM adaptively designs training cases based on real-time feedback from the evolutionary optimization process. We investigate how different feedback modalities, ranging from numeric metrics alone to combinations with plots and behavior visualizations, influence the LLM ability to generate meaningful curricula. Through a 2D robot navigation case study, tackled with genetic programming as optimizer, we evaluate our approach against static LLM-generated curricula and expert-designed baselines. We show that interactive curriculum generation outperforms static approaches, with multimodal feedback incorporating both progression plots and behavior visualizations yielding performance competitive with expert-designed curricula. This work contributes to understanding how LLMs can serve as interactive curriculum designers for embodied AI systems, with potential extensions to broader evolutionary robotics applications.
Authors: Kazuki Ota, Takayuki Osa, Motoki Omura, Tatsuya Harada
Abstract: Board games have long served as complex decision-making benchmarks in artificial intelligence. In this field, search-based reinforcement learning methods such as AlphaZero have achieved remarkable success. However, their significant computational demands have been pointed out as barriers to their reproducibility. In this study, we propose a model-free reinforcement learning algorithm designed for board games to achieve more efficient learning. To validate the efficiency of the proposed method, we conducted comprehensive experiments on five board games: Animal Shogi, Gardner Chess, Go, Hex, and Othello. The results demonstrate that the proposed method achieves more efficient learning than existing methods across these environments. In addition, our extensive ablation study shows the importance of core techniques used in the proposed method. We believe that our efficient algorithm shows the potential of model-free reinforcement learning in domains traditionally dominated by search-based methods.
Authors: Zhenhua Zou, Sheng Guo, Qiuyang Zhan, Lepeng Zhao, Shuo Li, Qi Li, Ke Xu, Mingwei Xu, Zhuotao Liu
Abstract: The evolution of Large Language Models (LLMs) has shifted mobile computing from App-centric interactions to system-level autonomous agents. Current implementations predominantly rely on a "Screen-as-Interface" paradigm, which inherits structural vulnerabilities and conflicts with the mobile ecosystem's economic foundations. In this paper, we conduct a systematic security analysis of state-of-the-art mobile agents using Doubao Mobile Assistant as a representative case. We decompose the threat landscape into four dimensions - Agent Identity, External Interface, Internal Reasoning, and Action Execution - revealing critical flaws such as fake App identity, visual spoofing, indirect prompt injection, and unauthorized privilege escalation stemming from a reliance on unstructured visual data. To address these challenges, we propose Aura, an Agent Universal Runtime Architecture for a clean-slate secure agent OS. Aura replaces brittle GUI scraping with a structured, agent-native interaction model. It adopts a Hub-and-Spoke topology where a privileged System Agent orchestrates intent, sandboxed App Agents execute domain-specific tasks, and the Agent Kernel mediates all communication. The Agent Kernel enforces four defense pillars: (i) cryptographic identity binding via a Global Agent Registry; (ii) semantic input sanitization through a multilayer Semantic Firewall; (iii) cognitive integrity via taint-aware memory and plan-trajectory alignment; and (iv) granular access control with non-deniable auditing. Evaluation on MobileSafetyBench shows that, compared to Doubao, Aura improves low-risk Task Success Rate from roughly 75% to 94.3%, reduces high-risk Attack Success Rate from roughly 40% to 4.4%, and achieves near-order-of-magnitude latency gains. These results demonstrate Aura as a viable, secure alternative to the "Screen-as-Interface" paradigm.
Authors: Rashid Mushkani
Abstract: Participatory approaches are widely invoked in AI governance, yet participation rarely translates into durable influence. In public sector and civic AI systems, community contributions such as deliberations, annotations, prompts, and incident reports are often recorded informally, weakly linked to system updates, and disconnected from enforceable rights or sustained compensation. As a result, participation is frequently symbolic rather than accountable. We introduce the Participation Ledger, a machine readable and auditable framework that operationalizes participation as traceable influence, enforceable authority, and compensable labor. The ledger represents participation as an influence graph that links contributed artifacts to verified changes in AI systems, including datasets, prompts, adapters, policies, guardrails, and evaluation suites. It integrates three elements: a Participation Evidence Standard documenting consent, privacy, compensation, and reuse terms; an influence tracing mechanism that connects system updates to replayable before and after tests, enabling longitudinal monitoring of commitments; and encoded rights and incentives. Capability Vouchers allow authorized community stewards to request or constrain specific system capabilities within defined boundaries, while Participation Credits support ongoing recognition and compensation when contributed tests continue to provide value. We ground the framework in four urban AI and public space governance deployments and provide a machine readable schema, templates, and an evaluation plan for assessing traceability, enforceability, and compensation in practice.
Authors: Bijean Ghafouri, Dorsaf Sallami, Luca Luceri, Taylor Lynn Curtis, Jean-Francois Godbout, Emilio Ferrara, Reihaneh Rabbany
Abstract: Research on misinformation has focused almost exclusively on supply, asking what falsehoods circulate, who produces them, and whether corrections work. A basic demand-side question remains unanswered. When ordinary people can fact-check anything they want, what do they actually ask about? We provide the first large-scale evidence on this question by analyzing close to 2{,}500 statements submitted by 457 participants to an open-ended AI fact-checking system. Each claim is classified along five semantic dimensions (domain, epistemic form, verifiability, target entity, and temporal reference), producing a behavioral map of public verification demand. Three findings stand out. First, users range widely across topics but default to a narrow epistemic repertoire, overwhelmingly submitting simple descriptive claims about present-day observables. Second, roughly one in four requests concerns statements that cannot be empirically resolved, including moral judgments, speculative predictions, and subjective evaluations, revealing a systematic mismatch between what users seek from fact-checking tools and what such tools can deliver. Third, comparison with the FEVER benchmark dataset exposes sharp structural divergences across all five dimensions, indicating that standard evaluation corpora encode a synthetic claim environment that does not resemble real-world verification needs. These results reframe fact-checking as a demand-driven problem and identify where current AI systems and benchmarks are misaligned with the uncertainty people actually experience.
Authors: Kacper Dudzic, Karolina Dro\.zd\.z, Maciej Wodzi\'nski, Anastazja Szu{\l}a, Marcin Moskalewicz
Abstract: Disturbances in temporality, such as desynchronization with the social environment and its unpredictability, are considered core features of autism with a deep impact on relationships. However, limitations regarding research on this issue include: 1) the dominance of deficit-based medical models of autism, 2) sample size in qualitative research, and 3) the lack of phenomenological anchoring in computational research. To bridge the gap between phenomenological and computational approaches and overcome sample-size limitations, our research integrated three methodologies. Study A: structured phenomenological interviews with autistic individuals using the Transdiagnostic Assessment of Temporal Experience. Study B: computational analysis of an autobiographical corpus of autistic narratives built for this purpose. Study C: a replication of a computational study using narrative flow measures to assess the perceived phenomenological authenticity of autistic autobiographies. Interviews revealed that the most significant differences between the autistic and control groups concerned unpredictability of experience. Computational results mirrored these findings: the temporal lexicon in autistic narratives was significantly more negatively valenced - particularly the "Immediacy & Suddenness" category. Outlier analysis identified terms associated with perceived discontinuity (unpredictably, precipitously, and abruptly) as highly negative. The computational analysis of narrative flow found that the autistic narratives contained within the corpus quantifiably resemble autobiographical stories more than imaginary ones. Overall, the temporal challenges experienced by autistic individuals were shown to primarily concern lived unpredictability and stem from the contents of lived experience, and not from autistic narrative construction.
Authors: Mingyu Cao, Alvaro Correia, Christos Louizos, Shiwei Liu, Lu Yin
Abstract: Diffusion Language Models (DLMs) generate text by iteratively denoising a masked sequence, repeatedly deciding which positions to commit at each step. Standard decoding follows a greedy rule: unmask the most confident positions, yet this local choice can lock the model into a suboptimal unmasking order, especially on reasoning-heavy prompts. We present SOAR, a training-free decoding algorithm that adapts its behavior to the model's uncertainty. When confidence is low, SOAR briefly widens the search over alternative unmasking decisions to avoid premature commitments; when confidence is high, it collapses the search and decodes many positions in parallel to reduce the number of denoising iterations. Across mathematical reasoning and code generation benchmarks (GSM8K, MBPP, HumanEval) on Dream-7B and LLaDA-8B, SOAR improves generation quality while maintaining competitive inference speed, offering a practical way to balance quality and efficiency in DLM decoding.
Authors: Feilong Liu
Abstract: Rotary positional embeddings (RoPE) are widely used in large language models to encode token positions through multiplicative rotations, yet their behavior at long context lengths remains poorly characterized. In this work, we reinterpret RoPE as phase modulation applied to a bank of complex oscillators, enabling analysis through classical signal processing theory. Under this formulation, we derive principled lower bounds on the RoPE base parameter that are necessary to preserve positional coherence over a target context length. These include a fundamental aliasing bound, analogous to a Nyquist limit, and a DC-component stability bound that constrains phase drift in low-frequency positional modes. We further extend this analysis to deep transformers, showing that repeated rotary modulation across layers compounds angular misalignment, tightening the base requirement as depth increases. Complementing these results, we derive a precision-dependent upper bound on the RoPE base arising from finite floating-point resolution. Beyond this limit, incremental phase updates become numerically indistinguishable, leading to positional erasure even in the absence of aliasing. Together, the lower and upper bounds define a precision- and depth-dependent feasibility region a Goldilocks zone for long-context transformers. We validate the framework through a comprehensive case study of state-of-the-art models, including LLaMA, Mistral, and DeepSeek variants, showing that observed successes, failures, and community retrofits align closely with the predicted bounds. Notably, models that violate the stability bound exhibit attention collapse and long-range degradation, while attempts to scale beyond one million tokens encounter a hard precision wall independent of architecture or training.
Authors: Samanta Ghosh, Shaila Afroz Anika, Umma Habiba Ahmed, B. M. Shahria Alam, Mohammad Tahmid Noor, Nishat Tasnim Niloy
Abstract: Guava fruits often suffer from many diseases. This can harm fruit quality and fruit crop yield. Early identification is important for minimizing damage and ensuring fruit health. This study focuses on 3 different categories for classifying diseases. These are Anthracnose, Fruit flies, and Healthy fruit. The data set used in this study is collected from Mendeley Data. This dataset contains 473 original images of Guava. These images vary in size and format. The original dataset was resized to 256x256 pixels with RGB color mode for better consistency. After this, the Data augmentation process is applied to improve the dataset by generating variations of the original images. The augmented dataset consists of 3784 images using advanced preprocessing techniques. Two deep learning models were implemented to classify the images. The InceptionV3 model is well known for its advanced framework. These apply multiple convolutional filters for obtaining different features effectively. On the other hand, the ResNet50 model helps to train deeper networks by using residual learning. The InceptionV3 model achieved the impressive accuracy of 98.15%, and ResNet50got 94.46% accuracy. Data mixing methods such as CutMix and MixUp were applied to enhance the model's robustness. The confusion matrix was used to evaluate the overall model performance of both InceptionV3 and Resnet50. Additionally, SHAP analysis is used to improve interpretability, which helps to find the significant parts of the image for the model prediction. This study purposes to highlight how advanced models enhan
Authors: Qixing Zhou, Jiacheng Zhang, Haiyang Wang, Rui Hao, Jiahe Wang, Minghao Han, Yuxue Yang, Shuzhe Wu, Feiyang Pan, Lue Fan, Dandan Tu, Zhaoxiang Zhang
Abstract: Agents powered by large language models (LLMs) are increasingly adopted in the software industry, contributing code as collaborators or even autonomous developers. As their presence grows, it becomes important to assess the current boundaries of their coding abilities. Existing agentic coding benchmarks, however, cover a limited task scope, e.g., bug fixing within a single pull request (PR), and often rely on non-executable evaluations or lack an automated approach for continually updating the evaluation coverage. To address such issues, we propose FeatureBench, a benchmark designed to evaluate agentic coding performance in end-to-end, feature-oriented software development. FeatureBench incorporates an execution-based evaluation protocol and a scalable test-driven method that automatically derives tasks from code repositories with minimal human effort. By tracing from unit tests along a dependency graph, our approach can identify feature-level coding tasks spanning multiple commits and PRs scattered across the development timeline, while ensuring the proper functioning of other features after the separation. Using this framework, we curated 200 challenging evaluation tasks and 3825 executable environments from 24 open-source repositories in the first version of our benchmark. Empirical evaluation reveals that the state-of-the-art agentic model, such as Claude 4.5 Opus, which achieves a 74.4% resolved rate on SWE-bench, succeeds on only 11.0% of tasks, opening new opportunities for advancing agentic coding. Moreover, benefiting from our automated task collection toolkit, FeatureBench can be easily scaled and updated over time to mitigate data leakage. The inherent verifiability of constructed environments also makes our method potentially valuable for agent training.
Authors: Li Sun, Qiqi Wan, Suyang Zhou, Zhenhao Huang, Philip S. Yu
Abstract: Graphs are ubiquitous, and learning on graphs has become a cornerstone in artificial intelligence and data mining communities. Unlike pixel grids in images or sequential structures in language, graphs exhibit a typical non-Euclidean structure with complex interactions among the objects. This paper argues that Riemannian geometry provides a principled and necessary foundation for graph representation learning, and that Riemannian graph learning should be viewed as a unifying paradigm rather than a collection of isolated techniques. While recent studies have explored the integration of graph learning and Riemannian geometry, most existing approaches are limited to a narrow class of manifolds, particularly hyperbolic spaces, and often adopt extrinsic manifold formulations. We contend that the central mission of Riemannian graph learning is to endow graph neural networks with intrinsic manifold structures, which remains underexplored. To advance this perspective, we identify key conceptual and methodological gaps in existing approaches and outline a structured research agenda along three dimensions: manifold type, neural architecture, and learning paradigm. We further discuss open challenges, theoretical foundations, and promising directions that are critical for unlocking the full potential of Riemannian graph learning. This paper aims to provide a coherent viewpoint and to stimulate broader exploration of Riemannian geometry as a foundational framework for future graph learning research.
Authors: Ivan Vuli\'c, Adam Grycner, Quentin de Laroussilhe, Jonas Pfeiffer
Abstract: Despite its huge number of variants, standard Low-Rank Adaptation (LoRA) is still a dominant technique for parameter-efficient fine-tuning (PEFT). Nonetheless, it faces persistent challenges, including the pre-selection of an optimal rank and rank-specific hyper-parameters, as well as the deployment complexity of heterogeneous-rank modules and more sophisticated LoRA derivatives. In this work, we introduce LoRA-Squeeze, a simple and efficient methodology that aims to improve standard LoRA learning by changing LoRA module ranks either post-hoc or dynamically during training}. Our approach posits that it is better to first learn an expressive, higher-rank solution and then compress it, rather than learning a constrained, low-rank solution directly. The method involves fine-tuning with a deliberately high(er) source rank, reconstructing or efficiently approximating the reconstruction of the full weight update matrix, and then using Randomized Singular Value Decomposition (RSVD) to create a new, compressed LoRA module at a lower target rank. Extensive experiments across 13 text and 10 vision-language tasks show that post-hoc compression often produces lower-rank adapters that outperform those trained directly at the target rank, especially if a small number of fine-tuning steps at the target rank is allowed. Moreover, a gradual, in-tuning rank annealing variant of LoRA-Squeeze consistently achieves the best LoRA size-performance trade-off.
Authors: Ali Tehrani, Yahya Emara, Essam Wissam, Wojciech Paluch, Waleed Atallah, {\L}ukasz Dudziak, Mohamed S. Abdelfattah
Abstract: Developing efficient GPU kernels is essential for scaling modern AI systems, yet it remains a complex task due to intricate hardware architectures and the need for specialized optimization expertise. Although Large Language Models (LLMs) demonstrate strong capabilities in general sequential code generation, they face significant challenges in GPU code generation because of the scarcity of high-quality labeled training data, compiler biases when generating synthetic solutions, and limited generalization across hardware generations. This precludes supervised fine-tuning (SFT) as a scalable methodology for improving current LLMs. In contrast, reinforcement learning (RL) offers a data-efficient and adaptive alternative but requires access to relevant tools, careful selection of training problems, and a robust evaluation environment. We present Makora's environment and tools for reinforcement learning finetuning of frontier models and report our results from fine-tuning GPT-5 for Triton code generation. In the single-attempt setting, our fine-tuned model improves kernel correctness from 43.7% to 77.0% (+33.3 percentage points) and increases the fraction of problems outperforming TorchInductor from 14.8% to 21.8% (+7 percentage points) compared to baseline GPT-5, while exceeding prior state-of-the-art models on KernelBench. When integrated into a full coding agent, it is able to solve up to 97.4% of problems in an expanded KernelBench suite, outperforming the PyTorch TorchInductor compiler on 72.9% of problems with a geometric mean speedup of 2.12x. Our work demonstrates that targeted post-training with reinforcement learning can unlock LLM capabilities in highly specialized technical domains where traditional supervised learning is limited by data availability, opening new pathways for AI-assisted accelerator programming.
Authors: Liangkai Liu, Kang G. Shin, Jinkyu Lee, Chengmo Yang, Weisong Shi
Abstract: Autonomous vehicles (AVs) rely on sensors and deep neural networks (DNNs) to perceive their surrounding environment and make maneuver decisions in real time. However, achieving real-time DNN inference in the AV's perception pipeline is challenging due to the large gap between the computation requirement and the AV's limited resources. Most, if not all, of existing studies focus on optimizing the DNN inference time to achieve faster perception by compressing the DNN model with pruning and quantization. In contrast, we present a Predictable Perception system with DNNs (PP-DNN) that reduce the amount of image data to be processed while maintaining the same level of accuracy for multi-tenant DNNs by dynamically selecting critical frames and regions of interest (ROIs). PP-DNN is based on our key insight that critical frames and ROIs for AVs vary with the AV's surrounding environment. However, it is challenging to identify and use critical frames and ROIs in multi-tenant DNNs for predictable inference. Given image-frame streams, PP-DNN leverages an ROI generator to identify critical frames and ROIs based on the similarities of consecutive frames and traffic scenarios. PP-DNN then leverages a FLOPs predictor to predict multiply-accumulate operations (MACs) from the dynamic critical frames and ROIs. The ROI scheduler coordinates the processing of critical frames and ROIs with multiple DNN models. Finally, we design a detection predictor for the perception of non-critical frames. We have implemented PP-DNN in an ROS-based AV pipeline and evaluated it with the BDD100K and the nuScenes dataset. PP-DNN is observed to significantly enhance perception predictability, increasing the number of fusion frames by up to 7.3x, reducing the fusion delay by >2.6x and fusion-delay variations by >2.3x, improving detection completeness by 75.4% and the cost-effectiveness by up to 98% over the baseline.
Authors: Ammar Ali, Baher Mohammad, Denis Makhov, Dmitriy Shopkhoev, Magauiya Zhussip, Stamatios Lefkimmiatis
Abstract: We present ROCKET, a training-free model compression method that achieves state-of-the-art performance in comparison with factorization, structured-sparsification and dynamic compression baselines. Operating under a global compression budget, ROCKET comprises two key innovations: First, it formulates layer-wise compression allocation as a multi-choice knapsack problem, selecting the optimal compression level for each layer to minimize total reconstruction error while adhering to a target model size. Second, it introduces a single-step sparse matrix factorization inspired by dictionary learning: using only a small calibration set, it sparsifies weight coefficients based on activation-weights sensitivity and then updates the dictionary in closed form via least squares bypassing iterative optimization, sparse coding, or backpropagation entirely. ROCKET consistently outperforms existing compression approaches across different model architectures at 20-50\% compression rates. Notably, it retains over 90\% of the original model's performance at 30\% compression without any fine-tuning. Moreover, when applying a light fine-tuning phase, recovery is substantially enhanced: for instance, compressing Qwen3-14B to an 8B-parameter model and healing it with just 30 million tokens yields performance nearly on par with the original Qwen3-8B. The code for ROCKET is at github.com/mts-ai/ROCKET/tree/main.
Authors: Valery Khvatov, Alexey Neyman
Abstract: Formal privacy metrics provide compliance-oriented guarantees but often fail to quantify actual linkability in released datasets. We introduce CVPL (Cluster-Vector-Projection Linkage), a geometric framework for post-hoc assessment of linkage risk between original and protected tabular data. CVPL represents linkage analysis as an operator pipeline comprising blocking, vectorization, latent projection, and similarity evaluation, yielding continuous, scenario-dependent risk estimates rather than binary compliance verdicts. We formally define CVPL under an explicit threat model and introduce threshold-aware risk surfaces, R(lambda, tau), that capture the joint effects of protection strength and attacker strictness. We establish a progressive blocking strategy with monotonicity guarantees, enabling anytime risk estimation with valid lower bounds. We demonstrate that the classical Fellegi-Sunter linkage emerges as a special case of CVPL under restrictive assumptions, and that violations of these assumptions can lead to systematic over-linking bias. Empirical validation on 10,000 records across 19 protection configurations demonstrates that formal k-anonymity compliance may coexist with substantial empirical linkability, with a significant portion arising from non-quasi-identifier behavioral patterns. CVPL provides interpretable diagnostics identifying which features drive linkage feasibility, supporting privacy impact assessment, protection mechanism comparison, and utility-risk trade-off analysis.
Authors: Jinxin Yu, Yudong Pan, Mengdi Wang, Huawei Li, Yinhe Han, Xiaowei Li, Ying Wang
Abstract: Transformer-based models dominate modern AI workloads but exacerbate memory bottlenecks due to their quadratic attention complexity and ever-growing model sizes. Existing accelerators, such as Groq and Cerebras, mitigate off-chip traffic with large on-chip caches, while algorithmic innovations such as FlashAttention fuse operators to avoid materializing large attention matrices. However, as off-chip traffic decreases, our measurements show that on-chip SRAM accesses account for over 60% of energy in long-sequence workloads, making cache access the new bottleneck. We propose 3D-Flow, a hybrid-bonded, 3D-stacked spatial accelerator that enables register-to-register communication across vertically partitioned PE tiers. Unlike 2D multi-array architectures limited by NoC-based router-to-router transfers, 3D-Flow leverages sub-10 um vertical TSVs to sustain cycle-level operator pipelining with minimal overhead. On top of this architecture, we design 3D-FlashAttention, a fine-grained scheduling method that balances latency across tiers, forming a bubble-free vertical dataflow without on-chip SRAM roundtrips. Evaluations on Transformer workloads (OPT and QWEN models) show that our 3D spatial accelerator reduces 46-93% energy consumption and achieves 1.4x-7.6x speedups compared to state-of-the-art 2D and 3D designs.
Authors: Returaj Burnwal, Nirav Pravinbhai Bhatt, Balaraman Ravindran
Abstract: This work addresses the problem of offline safe imitation learning (IL), where the goal is to learn safe and reward-maximizing policies from demonstrations that do not have per-timestep safety cost or reward information. In many real-world domains, online learning in the environment can be risky, and specifying accurate safety costs can be difficult. However, it is often feasible to collect trajectories that reflect undesirable or unsafe behavior, implicitly conveying what the agent should avoid. We refer to these as non-preferred trajectories. We propose a novel offline safe IL algorithm, OSIL, that infers safety from non-preferred demonstrations. We formulate safe policy learning as a Constrained Markov Decision Process (CMDP). Instead of relying on explicit safety cost and reward annotations, OSIL reformulates the CMDP problem by deriving a lower bound on reward maximizing objective and learning a cost model that estimates the likelihood of non-preferred behavior. Our approach allows agents to learn safe and reward-maximizing behavior entirely from offline demonstrations. We empirically demonstrate that our approach can learn safer policies that satisfy cost constraints without degrading the reward performance, thus outperforming several baselines.
Authors: Meizhong Wang, Wanxin Jin, Kun Cao, Lihua Xie, Yiguang Hong
Abstract: Developing world models that understand complex physical interactions is essential for advancing robotic planning and simulation.However, existing methods often struggle to accurately model the environment under conditions of data scarcity and complex contact-rich dynamic motion.To address these challenges, we propose ContactGaussian-WM, a differentiable physics-grounded rigid-body world model capable of learning intricate physical laws directly from sparse and contact-rich video sequences.Our framework consists of two core components: (1) a unified Gaussian representation for both visual appearance and collision geometry, and (2) an end-to-end differentiable learning framework that differentiates through a closed-form physics engine to infer physical properties from sparse visual observations.Extensive simulations and real-world evaluations demonstrate that ContactGaussian-WM outperforms state-of-the-art methods in learning complex scenarios, exhibiting robust generalization capabilities.Furthermore, we showcase the practical utility of our framework in downstream applications, including data synthesis and real-time MPC.
Authors: Rishikesh Bhyri, Brian R Quaranto, Philip J Seger, Kaity Tung, Brendan Fox, Gene Yang, Steven D. Schwaitzberg, Junsong Yuan, Nan Xi, Peter C W Kim
Abstract: Accurate counting of surgical instruments in Operating Rooms (OR) is a critical prerequisite for ensuring patient safety during surgery. Despite recent progress of large visual-language models and agentic AI, accurately counting such instruments remains highly challenging, particularly in dense scenarios where instruments are tightly clustered. To address this problem, we introduce Chain-of-Look, a novel visual reasoning framework that mimics the sequential human counting process by enforcing a structured visual chain, rather than relying on classic object detection which is unordered. This visual chain guides the model to count along a coherent spatial trajectory, improving accuracy in complex scenes. To further enforce the physical plausibility of the visual chain, we introduce the neighboring loss function, which explicitly models the spatial constraints inherent to densely packed surgical instruments. We also present SurgCount-HD, a new dataset comprising 1,464 high-density surgical instrument images. Extensive experiments demonstrate that our method outperforms state-of-the-art approaches for counting (e.g., CountGD, REC) as well as Multimodality Large Language Models (e.g., Qwen, ChatGPT) in the challenging task of dense surgical instrument counting.
Authors: Artsvik Avetisyan, Sachin Kumar
Abstract: Background: Subtle changes in spontaneous language production are among the earliest indicators of cognitive decline. Identifying linguistically interpretable markers of dementia can support transparent and clinically grounded screening approaches. Methods: This study analyzes spontaneous speech transcripts from the DementiaBank Pitt Corpus using three linguistic representations: raw cleaned text, a part-of-speech (POS)-enhanced representation combining lexical and grammatical information, and a POS-only syntactic representation. Logistic regression and random forest models were evaluated under two protocols: transcript-level train-test splits and subject-level five-fold cross-validation to prevent speaker overlap. Model interpretability was examined using global feature importance, and statistical validation was conducted using Mann-Whitney U tests with Cliff's delta effect sizes. Results: Across representations, models achieved stable performance, with syntactic and grammatical features retaining strong discriminative power even in the absence of lexical content. Subject-level evaluation yielded more conservative but consistent results, particularly for POS-enhanced and POS-only representations. Statistical analysis revealed significant group differences in functional word usage, lexical diversity, sentence structure, and discourse coherence, aligning closely with machine learning feature importance findings. Conclusion: The results demonstrate that abstract linguistic features capture robust markers of early cognitive decline under clinically realistic evaluation. By combining interpretable machine learning with non-parametric statistical validation, this study supports the use of linguistically grounded features for transparent and reliable language-based cognitive screening.
Authors: Kevin Yandoka Denamgana\"i, Kartic Subr
Abstract: Despite emerging research on Language Models (LM), few approaches analyse the invertibility of LMs. That is, given a LM and a desirable target output sequence of tokens, determining what input prompts would yield the target output remains an open problem. We formulate this problem as a classical gradient-based optimisation. First, we propose a simple algorithm to achieve end-to-end differentiability of a given (frozen) LM and then find optimised prompts via gradient descent. Our central insight is to view LMs as functions operating on sequences of distributions over tokens (rather than the traditional view as functions on sequences of tokens). Our experiments and ablations demonstrate that our DLM-powered inversion can reliably and efficiently optimise prompts of lengths $10$ and $80$ for targets of length $20$, for several white-box LMs (out-of-the-box).
Authors: Maciej Besta, {\L}ukasz Jarmocik, Orest Hrycyna, Shachar Klaiman, Konrad M\k{a}czka, Robert Gerstenberger, J\"urgen M\"uller, Piotr Nyczyk, Hubert Niewiadomski, Torsten Hoefler
Abstract: Graphs are foundational across domains but remain hard to use without deep expertise. LLMs promise accessible natural language (NL) graph analytics, yet they fail to process industry-scale property graphs effectively and efficiently: such datasets are large, highly heterogeneous, structurally complex, and evolve dynamically. To address this, we devise a novel abstraction for complex multi-query analytics over such graphs. Its key idea is to replace brittle generation of graph queries directly from NL with planning over a Semantic Catalog that describes both the graph schema and the graph operations. Concretely, this induces a clean separation between a Semantic Plane for LLM planning and broader reasoning, and an Execution Plane for deterministic, database-grade query execution over the full dataset and tool implementations. This design yields substantial gains in both token efficiency and task effectiveness even with small-context LLMs. We use this abstraction as the basis of the first LLM-enhanced graph analytics framework called GraphSeek. GraphSeek achieves substantially higher success rates (e.g., 86% over enhanced LangChain) and points toward the next generation of affordable and accessible graph analytics that unify LLM reasoning with database-grade execution over large and complex property graphs.
Authors: Dingkun Zhou, Shuchang Pan, Jiachen Lian, Siddharth Banerjee, Sarika Pasumarthy, Dhruv Hebbar, Siddhant Patel, Zeyi Austin Li, Kan Jen Cheng, Sanay Bordia, Krish Patel, Akshaj Gupta, Tingle Li, Gopala Anumanchipalli
Abstract: Human conversation is organized by an implicit chain of thoughts that manifests as timed speech acts. Capturing this perceptual pathway is key to building natural full-duplex interactive systems. We introduce a framework that models this process as multi-level perception, and then reasons over conversational behaviors via a Graph-of-Thoughts (GoT). Our approach formalizes the intent-to-action pathway with a hierarchical labeling scheme, predicting high-level communicative intents and low-level speech acts to learn their causal and temporal dependencies. To train this system, we develop a high quality corpus that pairs controllable, event-rich dialogue data with human-annotated labels. The GoT framework structures streaming predictions as an evolving graph, enabling a transformer to forecast the next speech act, generate concise justifications for its decisions, and dynamically refine its reasoning. Experiments on both synthetic and real duplex dialogues show that the framework delivers robust behavior detection, produces interpretable reasoning chains, and establishes a foundation for benchmarking conversational reasoning in full duplex spoken dialogue systems.
Authors: Junfei Wu, Jian Guan, Qiang Liu, Shu Wu, Liang Wang, Wei Wu, Tienie Tan
Abstract: Current large vision-language models (LVLMs) typically rely on text-only reasoning based on a single-pass visual encoding, which often leads to loss of fine-grained visual information. Recently the proposal of ''thinking with images'' attempts to alleviate this limitation by manipulating images via external tools or code; however, the resulting visual states are often insufficiently grounded in linguistic semantics, impairing effective cross-modal alignment - particularly when visual semantics or geometric relationships must be reasoned over across distant regions or multiple images. To address these challenges, we propose ''chatting with images'', a new framework that reframes visual manipulation as language-guided feature modulation. Under the guidance of expressive language prompts, the model dynamically performs joint re-encoding over multiple image regions, enabling tighter coupling between linguistic reasoning and visual state updates. We instantiate this paradigm in ViLaVT, a novel LVLM equipped with a dynamic vision encoder explicitly designed for such interactive visual reasoning, and trained it with a two-stage curriculum combining supervised fine-tuning and reinforcement learning to promote effective reasoning behaviors. Extensive experiments across eight benchmarks demonstrate that ViLaVT achieves strong and consistent improvements, with particularly pronounced gains on complex multi-image and video-based spatial reasoning tasks.
Authors: Kavan Fatehi, Mostafa Rahmani Ghourtani, Amir Sonee, Poonam Yadav, Alessandra M Russo, Hamed Ahmadi, Radu Calinescu
Abstract: Sixth-generation (6G) radio access networks (RANs) must enforce strict service-level agreements (SLAs) for heterogeneous slices, yet sudden latency spikes remain difficult to diagnose and resolve with conventional deep reinforcement learning (DRL) or explainable RL (XRL). We propose \emph{Attention-Enhanced Multi-Agent Proximal Policy Optimization (AE-MAPPO)}, which integrates six specialized attention mechanisms into multi-agent slice control and surfaces them as zero-cost, faithful explanations. The framework operates across O-RAN timescales with a three-phase strategy: predictive, reactive, and inter-slice optimization. A URLLC case study shows AE-MAPPO resolves a latency spike in $18$ms, restores latency to $0.98$ms with $99.9999\%$ reliability, and reduces troubleshooting time by $93\%$ while maintaining eMBB and mMTC continuity. These results confirm AE-MAPPO's ability to combine SLA compliance with inherent interpretability, enabling trustworthy and real-time automation for 6G RAN slicing.
Authors: Frank Xiao, Santiago Aranguri
Abstract: We propose activation-based data attribution, a method that traces behavioral changes in post-trained language models to responsible training datapoints. By computing activation-difference vectors for both test prompts and preference pairs and ranking by cosine similarity, we identify datapoints that cause specific behaviors and validate these attributions causally by retraining with modified data. Clustering behavior-datapoint similarity matrices also enables unsupervised discovery of emergent behaviors. Applying this to OLMo 2's production DPO training, we surfaced distractor-triggered compliance: a harmful behavior where the model complies with dangerous requests when benign formatting instructions are appended. Filtering top-ranked datapoints reduces this behavior by 63% while switching their labels achieves 78%. Our method outperforms gradient-based attribution and LLM-judge baselines while being over 10 times cheaper than both. This in-the-wild model organism - emerging from contaminated preference data rather than deliberate injection - provides a realistic benchmark for safety techniques.
Authors: Sebastian Wind, Jeta Sopa, Laurin Schmid, Quirin Jackl, Sebastian Kiefer, Fei Wu, Martin Mayr, Harald K\"ostler, Gerhard Wellein, Andreas Maier, Soroosh Tayebi Arasteh
Abstract: Large language models (LLMs) demonstrate strong general reasoning and language understanding, yet their performance degrades in domains governed by strict formal rules, precise terminology, and legally binding structure. Tax law exemplifies these challenges, as correct answers require exact statutory citation, structured legal argumentation, and numerical accuracy under rigid grading schemes. We algorithmically generate SteuerEx, the first open benchmark derived from authentic German university tax law examinations. SteuerEx comprises 115 expert-validated examination questions spanning six core tax law domains and multiple academic levels, and employs a statement-level, partial-credit evaluation framework that closely mirrors real examination practice. We further present SteuerLLM, a domain-adapted LLM for German tax law trained on a large-scale synthetic dataset generated from authentic examination material using a controlled retrieval-augmented pipeline. SteuerLLM (28B parameters) consistently outperforms general-purpose instruction-tuned models of comparable size and, in several cases, substantially larger systems, demonstrating that domain-specific data and architectural adaptation are more decisive than parameter scale for performance on realistic legal reasoning tasks. All benchmark data, training datasets, model weights, and evaluation code are released openly to support reproducible research in domain-specific legal artificial intelligence. A web-based demo of SteuerLLM is available at https://steuerllm.i5.ai.fau.de.
Authors: Yuheng Luo, Shuyan Li, Zhong Cao
Abstract: Feature selection remains a major challenge in medical prediction, where existing approaches such as LASSO often lack robustness and interpretability. We introduce GRASP, a novel framework that couples Shapley value driven attribution with group $L_{21}$ regularization to extract compact and non-redundant feature sets. GRASP first distills group level importance scores from a pretrained tree model via SHAP, then enforces structured sparsity through group $L_{21}$ regularized logistic regression, yielding stable and interpretable selections. Extensive comparisons with LASSO, SHAP, and deep learning based methods show that GRASP consistently delivers comparable or superior predictive accuracy, while identifying fewer, less redundant, and more stable features.
Authors: Jianxun Wang, Grant C. Forbes, Leonardo Villalobos-Arias, David L. Roberts
Abstract: Offline RL algorithms aim to improve upon the behavior policy that produces the collected data while constraining the learned policy to be within the support of the dataset. However, practical offline datasets often contain examples with little diversity or limited exploration of the environment, and from multiple behavior policies with diverse expertise levels. Limited exploration can impair the offline RL algorithm's ability to estimate \textit{Q} or \textit{V} values, while constraining towards diverse behavior policies can be overly conservative. Such datasets call for a balance between the RL objective and behavior policy constraints. We first identify the connection between $f$-divergence and optimization constraint on the Bellman residual through a more general Linear Programming form for RL and the convex conjugate. Following this, we introduce the general flexible function formulation for the $f$-divergence to incorporate an adaptive constraint on algorithms' learning objectives based on the offline training dataset. Results from experiments on the MuJoCo, Fetch, and AdroitHand environments show the correctness of the proposed LP form and the potential of the flexible $f$-divergence in improving performance for learning from a challenging dataset when applied to a compatible constrained optimization algorithm.
Authors: Yicheng Chen, Zerun Ma, Xinchen Xie, Yining Li, Kai Chen
Abstract: In the current landscape of Large Language Models (LLMs), the curation of large-scale, high-quality training data is a primary driver of model performance. A key lever is the \emph{data recipe}, which comprises a data processing pipeline to transform raw sources into training corpora. Despite the growing use of LLMs to automate individual data processing steps, such as data synthesis and filtering, the overall design of data recipes remains largely manual and labor-intensive, requiring substantial human expertise and iteration. To bridge this gap, we formulate \emph{end-to-end data recipe generation} for LLM adaptation. Given a target benchmark and a pool of available data sources, a model is required to output a complete data recipe that adapts a base LLM to the target task. We present DataChef-32B, which performs online reinforcement learning using a proxy reward that predicts downstream performance for candidate recipes. Across six held-out tasks, DataChef-32B produces practical recipes that reach comparable downstream performance to those curated by human experts. Notably, the recipe from DataChef-32B adapts Qwen3-1.7B-Base to the math domain, achieving 66.7 on AIME'25 and surpassing Qwen3-1.7B. This work sheds new light on automating LLM training and developing self-evolving AI systems.
Authors: Carlos Stein Brito
Abstract: Neural PDE surrogates are often deployed in data-limited or partially observed regimes where downstream decisions depend on calibrated uncertainty in addition to low prediction error. Existing approaches obtain uncertainty through ensemble replication, fixed stochastic noise such as dropout, or post hoc calibration. Cross-regularized uncertainty learns uncertainty parameters during training using gradients routed through a held-out regularization split. The predictor is optimized on the training split for fit, while low-dimensional uncertainty controls are optimized on the regularization split to reduce train-test mismatch, yielding regime-adaptive uncertainty without per-regime noise tuning. The framework can learn continuous noise levels at the output head, within hidden features, or within operator-specific components such as spectral modes. We instantiate the approach in Fourier Neural Operators and evaluate on APEBench sweeps over observed fraction and training-set size. Across these sweeps, the learned predictive distributions are better calibrated on held-out splits and the resulting uncertainty fields concentrate in high-error regions in one-step spatial diagnostics.
Authors: Soumya Suvra Ghosal, Souradip Chakraborty, Vaibhav Singh, Furong Huang, Dinesh Manocha, Amrit Singh Bedi
Abstract: Reinforcement learning (RL) based post-training for explicit chain-of-thought (e.g., GRPO) improves the reasoning ability of multimodal large-scale reasoning models (MLRMs). But recent evidence shows that it can simultaneously degrade safety alignment and increase jailbreak success rates. We propose SafeThink, a lightweight inference-time defense that treats safety recovery as a satisficing constraint rather than a maximization objective. SafeThink monitors the evolving reasoning trace with a safety reward model and conditionally injects an optimized short corrective prefix ("Wait, think safely") only when the safety threshold is violated. In our evaluations across six open-source MLRMs and four jailbreak benchmarks (JailbreakV-28K, Hades, FigStep, and MM-SafetyBench), SafeThink reduces attack success rates by 30-60% (e.g., LlamaV-o1: 63.33% to 5.74% on JailbreakV-28K, R1-Onevision: 69.07% to 5.65% on Hades) while preserving reasoning performance (MathVista accuracy: 65.20% to 65.00%). A key empirical finding from our experiments is that safety recovery is often only a few steering steps away: intervening in the first 1-3 reasoning steps typically suffices to redirect the full generation toward safe completions.
Authors: Jialiang Wang, Shengxiang Xu, Hanmo Liu, Jiachuan Wang, Yuyu Luo, Shimin Di, Min-Ling Zhang, Lei Chen
Abstract: Automatically generating agentic workflows -- executable operator graphs or codes that orchestrate reasoning, verification, and repair -- has become a practical way to solve complex tasks beyond what single-pass LLM generation can reliably handle. Yet what constitutes a good workflow depends heavily on the task distribution and the available operators. Under domain shift, current systems typically rely on iterative workflow refinement to discover a feasible workflow from a large workflow space, incurring high iteration costs and yielding unstable, domain-specific behavior. In response, we internalize a decompose-recompose-decide mechanism into an open-source LLM for cross-domain workflow generation. To decompose, we learn a compact set of reusable workflow capabilities across diverse domains. To recompose, we map each input task to a sparse composition over these bases to generate a task-specific workflow in a single pass. To decide, we attribute the success or failure of workflow generation to counterfactual contributions from learned capabilities, thereby capturing which capabilities actually drive success by their marginal effects. Across stringent multi-domain, cross-domain, and unseen-domain evaluations, our 1-pass generator surpasses SOTA refinement baselines that consume 20 iterations, while substantially reducing generation latency and cost.
Authors: Tessa Han, Sebastian Bordt, Hanlin Zhang, Sham Kakade
Abstract: The prevailing paradigm in large language model (LLM) development is to pretrain a base model, then perform further training to improve performance and model behavior. However, hyperparameter optimization and scaling laws have been studied primarily from the perspective of the base model's validation loss, ignoring downstream adaptability. In this work, we study pretraining from the perspective of model plasticity, that is, the ability of the base model to successfully adapt to downstream tasks through fine-tuning. We focus on the role of weight decay, a key regularization parameter during pretraining. Through systematic experiments, we show that models trained with larger weight decay values are more plastic, meaning they show larger performance gains when fine-tuned on downstream tasks. This phenomenon can lead to counterintuitive trade-offs where base models that perform worse after pretraining can perform better after fine-tuning. Further investigation of weight decay's mechanistic effects on model behavior reveals that it encourages linearly separable representations, regularizes attention matrices, and reduces overfitting on the training data. In conclusion, this work demonstrates the importance of using evaluation metrics beyond cross-entropy loss for hyperparameter optimization and casts light on the multifaceted role of that a single optimization hyperparameter plays in shaping model behavior.
Authors: Shaswat Garg, Matin Moezzi, Brandon Da Silva
Abstract: Hierarchical goal-conditioned reinforcement learning (H-GCRL) provides a powerful framework for tackling complex, long-horizon tasks by decomposing them into structured subgoals. However, its practical adoption is hindered by poor data efficiency and limited policy expressivity, especially in offline or data-scarce regimes. In this work, Normalizing flow-based hierarchical implicit Q-learning (NF-HIQL), a novel framework that replaces unimodal gaussian policies with expressive normalizing flow policies at both the high- and low-levels of the hierarchy is introduced. This design enables tractable log-likelihood computation, efficient sampling, and the ability to model rich multimodal behaviors. New theoretical guarantees are derived, including explicit KL-divergence bounds for Real-valued non-volume preserving (RealNVP) policies and PAC-style sample efficiency results, showing that NF-HIQL preserves stability while improving generalization. Empirically, NF-HIQL is evaluted across diverse long-horizon tasks in locomotion, ball-dribbling, and multi-step manipulation from OGBench. NF-HIQL consistently outperforms prior goal-conditioned and hierarchical baselines, demonstrating superior robustness under limited data and highlighting the potential of flow-based architectures for scalable, data-efficient hierarchical reinforcement learning.
Authors: Ruichuan An, Sihan Yang, Ziyu Guo, Wei Dai, Zijun Shen, Haodong Li, Renrui Zhang, Xinyu Wei, Guopeng Li, Wenshan Wu, Wentao Zhang
Abstract: Unified Multimodal Models (UMMs) have shown remarkable progress in visual generation. Yet, existing benchmarks predominantly assess $\textit{Crystallized Intelligence}$, which relies on recalling accumulated knowledge and learned schemas. This focus overlooks $\textit{Generative Fluid Intelligence (GFI)}$: the capacity to induce patterns, reason through constraints, and adapt to novel scenarios on the fly. To rigorously assess this capability, we introduce $\textbf{GENIUS}$ ($\textbf{GEN}$ Fluid $\textbf{I}$ntelligence Eval$\textbf{U}$ation $\textbf{S}$uite). We formalize $\textit{GFI}$ as a synthesis of three primitives. These include $\textit{Inducing Implicit Patterns}$ (e.g., inferring personalized visual preferences), $\textit{Executing Ad-hoc Constraints}$ (e.g., visualizing abstract metaphors), and $\textit{Adapting to Contextual Knowledge}$ (e.g., simulating counter-intuitive physics). Collectively, these primitives challenge models to solve problems grounded entirely in the immediate context. Our systematic evaluation of 12 representative models reveals significant performance deficits in these tasks. Crucially, our diagnostic analysis disentangles these failure modes. It demonstrates that deficits stem from limited context comprehension rather than insufficient intrinsic generative capability. To bridge this gap, we propose a training-free attention intervention strategy. Ultimately, $\textbf{GENIUS}$ establishes a rigorous standard for $\textit{GFI}$, guiding the field beyond knowledge utilization toward dynamic, general-purpose reasoning. Our dataset and code will be released at: $\href{https://github.com/arctanxarc/GENIUS}{https://github.com/arctanxarc/GENIUS}$.
URLs: https://github.com/arctanxarc/GENIUS, https://github.com/arctanxarc/GENIUS
Authors: Gongye Liu, Bo Yang, Yida Zhi, Zhizhou Zhong, Lei Ke, Didan Deng, Han Gao, Yongxiang Huang, Kaihao Zhang, Hongbo Fu, Wenhan Luo
Abstract: Preference optimization for diffusion and flow-matching models relies on reward functions that are both discriminatively robust and computationally efficient. Vision-Language Models (VLMs) have emerged as the primary reward provider, leveraging their rich multimodal priors to guide alignment. However, their computation and memory cost can be substantial, and optimizing a latent diffusion generator through a pixel-space reward introduces a domain mismatch that complicates alignment. In this paper, we propose DiNa-LRM, a diffusion-native latent reward model that formulates preference learning directly on noisy diffusion states. Our method introduces a noise-calibrated Thurstone likelihood with diffusion-noise-dependent uncertainty. DiNa-LRM leverages a pretrained latent diffusion backbone with a timestep-conditioned reward head, and supports inference-time noise ensembling, providing a diffusion-native mechanism for test-time scaling and robust rewarding. Across image alignment benchmarks, DiNa-LRM substantially outperforms existing diffusion-based reward baselines and achieves performance competitive with state-of-the-art VLMs at a fraction of the computational cost. In preference optimization, we demonstrate that DiNa-LRM improves preference optimization dynamics, enabling faster and more resource-efficient model alignment.
Authors: Manuel Mondal, Ljiljana Dolamic, G\'er\^ome Bovet, Philippe Cudr\'e-Mauroux, Julien Audiffren
Abstract: The handling of probabilities in the form of uncertainty or partial information is an essential task for LLMs in many settings and applications. A common approach to evaluate an LLM's probabilistic reasoning capabilities is to assess its ability to answer questions pertaining to probability through the use of multiple-choice questions (MCQs). However, this paradigm, which we refer to as explicit probabilistic reasoning, has been shown in the literature to yield significant limitations (e.g., sensitivity to answer ordering). In this work, we introduce an alternative approach, named implicit probabilistic reasoning, which evaluates the models' ability to integrate probabilistic reasoning into their text generation process. To achieve this, we rephrase MCQs as text-completion scenarios with a determined set of outcomes and compare the model's next-token probability assignments to the true likelihood of the outcomes. In line with previous work, we find that models exhibit solid performance in their explicit probabilistic reasoning (i.e., answers to MCQs). However, during text completion (i.e., implicit probabilistic reasoning), where the same information must be taken into account to generate text, the models' predictions often significantly diverge from the known ground truth. For instance, our evaluation method reveals that implicit probabilistic reasoning is improperly influenced by many factors, such as independent prior events, partial observations about a result, or statistical background information. All of these issues can cause erroneous results to be produced in text generation, which are not detected by conventional MCQ-based evaluation.
Authors: Prakhar Godara, Tilman Diego Alem\'an
Abstract: \textit{Reasoning} may be viewed as an algorithm $P$ that makes a choice of an action $a^* \in \mathcal{A}$, aiming to optimize some outcome. However, executing $P$ itself bears costs (time, energy, limited capacity, etc.) and needs to be considered alongside explicit utility obtained by making the choice in the underlying decision problem. Finding the right $P$ can itself be framed as an optimization problem over the space of reasoning processes $P$, generally referred to as \textit{metareasoning}. Conventionally, human metareasoning models assume that the agent knows the transition and reward distributions of the underlying MDP. This paper generalizes such models by proposing a meta Bayes-Adaptive MDP (meta-BAMDP) framework to handle metareasoning in environments with unknown reward/transition distributions, which encompasses a far larger and more realistic set of planning problems that humans and AI systems face. As a first step, we apply the framework to Bernoulli bandit tasks. Owing to the meta problem's complexity, our solutions are necessarily approximate. However, we introduce two novel theorems that significantly enhance the tractability of the problem, enabling stronger approximations that are robust within a range of assumptions grounded in realistic human decision-making scenarios. These results offer a resource-rational perspective and a normative framework for understanding human exploration under cognitive constraints, as well as providing experimentally testable predictions about human behavior in Bernoulli Bandit tasks.
Authors: Cristian Daniel P\u{a}duraru, Antonio B\u{a}rb\u{a}lau, Radu Filipescu, Andrei Liviu Nicolicioiu, Elena Burceanu
Abstract: Current methods for detecting spurious correlations rely on analyzing dataset statistics or error patterns, leaving many harmful shortcuts invisible when counterexamples are absent. We introduce BEE (Bridging Explainability and Embeddings), a framework that shifts the focus from model predictions to the weight space, and to the embedding geometry underlying decisions. By analyzing how fine-tuning perturbs pretrained representations, BEE uncovers spurious correlations that remain hidden from conventional evaluation pipelines. We use linear probing as a transparent diagnostic lens, revealing spurious features that not only persist after full fine-tuning but also transfer across diverse state-of-the-art models. Our experiments cover numerous datasets and domains: vision (Waterbirds, CelebA, ImageNet-1k), language (CivilComments, MIMIC-CXR medical notes), and multiple embedding families (CLIP, CLIP-DataComp.XL, mGTE, BLIP2, SigLIP2). BEE consistently exposes spurious correlations: from concepts that slash the ImageNet accuracy by up to 95%, to clinical shortcuts in MIMIC-CXR notes that induce dangerous false negatives. Together, these results position BEE as a general and principled tool for diagnosing spurious correlations in weight space, enabling principled dataset auditing and more trustworthy foundation models. The source code is publicly available at https://github.com/bit-ml/bee.
Authors: Abdelrhman Shaheen, Anas Badr, Ali Abohendy, Hatem Alsaadawy, Nadine Alsayad, Ehab H. El-Shazly
Abstract: Reinforcement Learning (RL) has been widely used in many applications, particularly in gaming, which serves as an excellent training ground for AI models. Google DeepMind has pioneered innovations in this field, employing reinforcement learning algorithms, including model-based, model-free, and deep Q-network approaches, to create advanced AI models such as AlphaGo, AlphaGo Zero, and MuZero. AlphaGo, the initial model, integrates supervised learning and reinforcement learning to master the game of Go, surpassing professional human players. AlphaGo Zero refines this approach by eliminating reliance on human gameplay data, instead utilizing self-play for enhanced learning efficiency. MuZero further extends these advancements by learning the underlying dynamics of game environments without explicit knowledge of the rules, achieving adaptability across various games, including complex Atari games. This paper reviews the significance of reinforcement learning applications in Atari and strategy-based games, analyzing these three models, their key innovations, training processes, challenges encountered, and improvements made. Additionally, we discuss advancements in the field of gaming, including MiniZero and multi-agent models, highlighting future directions and emerging AI models from Google DeepMind.
Authors: Haokun Liu, Sicong Huang, Jingyu Hu, Yangqiaoyu Zhou, Chenhao Tan
Abstract: There is growing interest in hypothesis generation with large language models (LLMs). However, fundamental questions remain: what makes a good hypothesis, and how can we systematically evaluate methods for hypothesis generation? To address this, we introduce HypoBench, a novel benchmark designed to evaluate LLMs and hypothesis generation methods across multiple aspects, including practical utility, generalizability, and hypothesis discovery rate. HypoBench includes 7 real-world tasks and 5 synthetic tasks with 194 distinct datasets. We evaluate four state-of-the-art LLMs combined with six existing hypothesis-generation methods. Overall, our results suggest that existing methods are capable of discovering valid and novel patterns in the data. However, the results from synthetic datasets indicate that there is still significant room for improvement, as current hypothesis generation methods do not fully uncover all relevant or meaningful patterns. Specifically, in synthetic settings, as task difficulty increases, performance significantly drops, with best models and methods only recovering 38.8% of the ground-truth hypotheses. These findings highlight challenges in hypothesis generation and demonstrate that HypoBench serves as a valuable resource for improving AI systems designed to assist scientific discovery.
Authors: Shuhang Xu, Weijian Deng, Yixuan Zhou, Fangwei Zhong
Abstract: Concepts serve as fundamental abstractions that support human reasoning and categorization. However, it remains unclear whether large language models truly capture such conceptual structures or primarily rely on surface-level pattern memorization. Existing benchmarks are largely static and fact oriented, which limits their ability to probe fine-grained semantic understanding and makes them vulnerable to data leakage and overfitting. To address this limitation, we introduce CK-Arena, a dynamic benchmark for conceptual knowledge evaluation based on a multi agent social deduction game, namely the Undercover game. In this setting, LLM based agents are assigned subtly different concept words and must describe, distinguish, and infer conceptual properties from others' statements. Model performance is evaluated through both game level outcomes and the semantic quality of generated descriptions. Furthermore, CK-Arena leverages the interaction process to automatically construct high quality question answering data for fine grained diagnostic analysis. Experimental results show that conceptual understanding varies substantially across models and categories, and is not strictly aligned with overall model capability. The data and code are available at the project homepage: https://ck-arena.site.
URLs: https://ck-arena.site.
Authors: Yisen Gao, Jiaxin Bai, Tianshi Zheng, Qingyun Sun, Ziwei Zhang, Xingcheng Fu, Jianxin Li, Yangqiu Song
Abstract: Abductive reasoning in knowledge graphs aims to generate plausible logical hypotheses from observed entities, with broad applications in areas such as clinical diagnosis and scientific discovery. However, due to a lack of controllability, a single observation may yield numerous plausible but redundant or irrelevant hypotheses on large-scale knowledge graphs. To address this limitation, we introduce the task of controllable hypothesis generation to improve the practical utility of abductive reasoning. This task faces two key challenges when controlling for generating long and complex logical hypotheses: hypothesis space collapse and hypothesis oversensitivity. To address these challenges, we propose CtrlHGen, a Controllable logcial Hypothesis Generation framework for abductive reasoning over knowledge graphs, trained in a two-stage paradigm including supervised learning and subsequent reinforcement learning. To mitigate hypothesis space collapse, we design a dataset augmentation strategy based on sub-logical decomposition, enabling the model to learn complex logical structures by leveraging semantic patterns in simpler components. To address hypothesis oversensitivity, we incorporate smoothed semantic rewards including Dice and Overlap scores, and introduce a condition-adherence reward to guide the generation toward user-specified control constraints. Extensive experiments on three benchmark datasets demonstrate that our model not only better adheres to control conditions but also achieves superior semantic similarity performance compared to baselines. Our code is available at https://github.com/HKUST-KnowComp/CtrlHGen.
Authors: Lintao Wang, Encheng Su, Jiaqi Liu, Pengze Li, Jiabei Xiao, Wenlong Zhang, Xinnan Dai, Xi Chen, Yuan Meng, Lei Bai, Wanli Ouyang, Shixiang Tang, Aoran Wang, Xinzhu Ma
Abstract: Physics problem-solving is a challenging domain for AI models, requiring integration of conceptual understanding, mathematical reasoning, and interpretation of physical diagrams. Existing evaluations fail to capture the full breadth and complexity of undergraduate physics, whereas this level provides a rigorous yet standardized testbed for pedagogical assessment of multi-step physical reasoning. To this end, we present PhysUniBench, a large-scale multimodal benchmark designed to evaluate and improve the reasoning capabilities of multimodal large language models (MLLMs) specifically on undergraduate-level physics problems. PhysUniBench consists of 3,304 physics questions spanning 8 major sub-disciplines of physics, each accompanied by one visual diagram. The benchmark includes both open-ended and multiple-choice questions, systematically curated and difficulty-rated through an iterative process. The benchmark's construction involved a rigorous multi-stage process, including multiple roll-outs, expert-level evaluation, automated filtering of easily solved problems, and a nuanced difficulty grading system with five levels. Through extensive experiments, we observe that current models encounter substantial challenges in physics reasoning, where GPT-5 achieves only 51.6% accuracy in the PhysUniBench. These results highlight that current MLLMs struggle with advanced physics reasoning, especially on multi-step problems and those requiring precise diagram interpretation. By providing a broad and rigorous assessment tool, PhysUniBench aims to drive progress in AI for Science, encouraging the development of models with stronger physical reasoning, problem-solving skills, and multimodal understanding.
Authors: Li Du, Hanyu Zhao, Yiming Ju, Tengfei Pan
Abstract: Instruction tuning has become a foundation for unlocking the capabilities of large-scale pretrained models and improving their performance on complex tasks. Thus, the construction of high-quality instruction datasets is crucial for enhancing model performance and generalizability. Although current instruction datasets have reached tens of millions of samples, models finetuned on them may still struggle with complex instruction following and tasks in rare domains. This is primarily due to limited expansion in both ``coverage'' (coverage of task types and knowledge areas) and ``depth'' (instruction complexity) of the instruction set. To address this issue, we propose a systematic instruction data construction framework, which integrates a hierarchical tagging system, an informative seed selection algorithm, an evolutionary data synthesis process, and a model deficiency diagnosis with targeted data generation. These components form an iterative closed-loop to continuously enhance the coverage and depth of instruction data. Based on this framework, we construct Infinity Instruct Subject, a high-quality dataset containing $\sim$1.5 million instructions. Experiments on multiple foundation models and benchmark tasks demonstrate its effectiveness in improving instruction-following capabilities. Further analyses suggest that Infinity Instruct Subject shows enlarged coverage and depth compared to comparable synthesized instruction datasets. Our work lays a theoretical and practical foundation for the efficient, continuous evolution of instruction datasets, moving from data quantity expansion to qualitative improvement.
Authors: Jackson Eshbaugh, Chetan Tiwari, Jorge Silveyra
Abstract: Computational models have emerged as powerful tools for energy modeling research, touting scalability and quantitative results. However, these models require a plethora of data, some of which can be inaccessible, expensive, or can raise privacy concerns. We introduce a modular multimodal framework to produce this data from publicly accessible images and residential information using generative Artificial Intelligence (AI). Additionally, we provide a pipeline demonstrating this framework and we evaluate its generative AI components. Our experiments show that our framework's use of AI avoids common issues with generative models and produces realistic multimodal data. By reducing dependence on costly or restricted data sources, we pave a path towards more accessible research in Machine Learning (ML) and other data-driven disciplines.
Authors: Xuemiao Zhang, Can Ren, Chengying Tu, Rongxiang Weng, Shuo Wang, Hongfei Yan, Jingang Wang, Xunliang Cai
Abstract: Recent progress in large reasoning models for challenging mathematical reasoning has been driven by reinforcement learning (RL). Incorporating long chain-of-thought (CoT) data during mid-training has also been shown to substantially improve reasoning depth. However, current approaches often utilize CoT data indiscriminately, leaving open the critical question of which data types most effectively enhance model reasoning capabilities. In this paper, we define the foundation model's reasoning potential for the first time as the inverse of the number of independent attempts required to correctly answer the question, which is strongly correlated with the final model performance. We then propose utilizing diverse data enriched with high-value reasoning patterns to expand the reasoning potential. Specifically, we abstract atomic reasoning patterns from CoT sequences, characterized by commonality and inductive capabilities, and use them to construct a core reference set enriched with valuable reasoning patterns. Furthermore, we propose a dual-granularity algorithm involving chains of reasoning patterns and token entropy, efficiently selecting high-value CoT data (CoTP) from the data pool that aligns with the core set, thereby training models to master reasoning effectively. Only 10B-token CoTP data enables the 85A6B Mixture-of-Experts (MoE) model to improve by 9.58% on the challenging AIME 2024 and 2025, and to raise the upper bound of downstream RL performance by 7.81%.
Authors: Yiquan Wang, Tin-Yeh Huang, Qingyun Gao, Yuhan Chang, Jialin Zhang
Abstract: Compound heatwaves increasingly trigger complex cascading failures that propagate through interconnected physical and human systems, yet the fragmentation of disciplinary knowledge hinders the comprehensive mapping of these systemic risk topologies. This study introduces the Heatwave Discovery Agent HeDA as an autonomous scientific synthesis framework designed to bridge cognitive gaps by constructing a high fidelity knowledge graph from 8,111 academic publications. By structuring 70,297 evidence nodes, the system exhibits enhanced inferential fidelity in capturing long tail risk mechanisms and achieves a significant accuracy margin compared to standard foundation models including GPT 5.2 and Claude Sonnet 4.5 in complex reasoning tasks. The resulting topological analysis reveals a critical bio ecological mediation effect where biological systems function as the primary non linear amplifiers of thermal stress that transform physical meteorological hazards into systemic socioeconomic losses. We further identify latent functional couplings between theoretically distinct sectors such as the heat induced synchronization of power grid failures and emergency medical capacity saturation. These findings elucidate the dynamics of compound climate risks and provide an empirical basis for shifting adaptation strategies from static sectoral defense to dynamic cross system resilience.
Authors: Yu Zeng, Wenxuan Huang, Shiting Huang, Xikun Bao, Yukun Qi, Yiming Zhao, Qiuchen Wang, Lin Chen, Zehui Chen, Huaian Chen, Wanli Ouyang, Feng Zhao
Abstract: Although current large Vision-Language Models (VLMs) have advanced in multimodal understanding and reasoning, their fundamental perceptual and reasoning abilities remain limited. Specifically, even on simple jigsaw tasks, existing VLMs perform near randomly, revealing deficiencies in core perception and reasoning capabilities. While high-quality vision-language data can enhance these capabilities, its scarcity and limited scalability impose significant constraints. To address this, we propose AGILE, an Agentic jiGsaw Interaction Learning for Enhancing visual perception and reasoning in VLMs. AGILE formulates jigsaw solving as an interactive process, enabling the model to progressively engage with the environment. At each step, the model generates executable code to perform an action based on the current state, while the environment provides fine-grained visual feedback to guide task completion. Through this iterative cycle of observation and interaction, the model incrementally improves its perceptual and reasoning capabilities via exploration and feedback. Experimental results show that AGILE not only substantially boosts performance on jigsaw tasks of varying complexity (e.g., increasing accuracy from 9.5% to 82.8% under the 2 $\times$ 2 setting) but also demonstrates strong generalization across 9 general vision tasks, achieving an average improvement of 3.1%. These results indicate notable enhancements in both perceptual and reasoning abilities. This work opens a new avenue for advancing reasoning and generalization in multimodal models and provides an efficient, scalable solution to the scarcity of multimodal reinforcement learning data. The code and datasets is available at https://github.com/yuzeng0-0/AGILE .
Authors: Rashid Mushkani
Abstract: Artificial intelligence systems increasingly mediate knowledge, communication, and decision making. Development and governance remain concentrated within a small set of firms and states, raising concerns that technologies may encode narrow interests and limit public agency. Capability benchmarks for language, vision, and coding are common, yet public, auditable measures of pluralistic governance are rare. We define AI pluralism as the degree to which affected stakeholders can shape objectives, data practices, safeguards, and deployment. We present the AI Pluralism Index (AIPI), a transparent, evidence-based instrument that evaluates producers and system families across four pillars: participatory governance, inclusivity and diversity, transparency, and accountability. AIPI codes verifiable practices from public artifacts and independent evaluations, explicitly handling "Unknown" evidence to report both lower-bound ("evidence") and known-only scores with coverage. We formalize the measurement model; implement a reproducible pipeline that integrates structured web and repository analysis, external assessments, and expert interviews; and assess reliability with inter-rater agreement, coverage reporting, cross-index correlations, and sensitivity analysis. The protocol, codebook, scoring scripts, and evidence graph are maintained openly with versioned releases and a public adjudication process. We report pilot provider results and situate AIPI relative to adjacent transparency, safety, and governance frameworks. The index aims to steer incentives toward pluralistic practice and to equip policymakers, procurers, and the public with comparable evidence.
Authors: Yisen Gao, Jiaxin Bai, Yi Huang, Xingcheng Fu, Qingyun Sun, Yangqiu Song
Abstract: Deductive and abductive reasoning are two critical paradigms for analyzing knowledge graphs, enabling applications from financial query answering to scientific discovery. Deductive reasoning on knowledge graphs usually involves retrieving entities that satisfy a complex logical query, while abductive reasoning generates plausible logical hypotheses from observations. Despite their clear synergistic potential, where deduction can validate hypotheses and abduction can uncover deeper logical patterns, existing methods address them in isolation. To bridge this gap, we propose DARK, a unified framework for Deductive and Abductive Reasoning in Knowledge graphs. As a masked diffusion model capable of capturing the bidirectional relationship between queries and conclusions, DARK has two key innovations. First, to better leverage deduction for hypothesis refinement during abductive reasoning, we introduce a self-reflective denoising process that iteratively generates and validates candidate hypotheses against the observed conclusion. Second, to discover richer logical associations, we propose a logic-exploration reinforcement learning approach that simultaneously masks queries and conclusions, enabling the model to explore novel reasoning compositions. Extensive experiments on multiple benchmark knowledge graphs show that DARK achieves state-of-the-art performance on both deductive and abductive reasoning tasks, demonstrating the significant benefits of our unified approach.
Authors: Deniz Gorur, Antonio Rago, Francesca Toni
Abstract: Judgmental forecasting is the task of making predictions about future events based on human judgment. This task can be seen as a form of claim verification, where the claim corresponds to a future event and the task is to assess the plausibility of that event. In this paper, we propose a novel multi-agent framework for claim verification, whereby different agents may disagree on claim veracity and bring specific evidence for and against the claims, represented as quantitative bipolar argumentation frameworks (QBAFs). We then instantiate the framework for supporting claim verification, with a variety of agents realised with Large Language Models (LLMs): (1) ArgLLM agents, an existing approach for claim verification that generates and evaluates QBAFs; (2) RbAM agents, whereby LLM-empowered Relation-based Argument Mining (RbAM) from external sources is used to generate QBAFs; (3) RAG-ArgLLM agents, extending ArgLLM agents with a form of Retrieval-Augmented Generation (RAG) of arguments from external sources. Finally, we conduct experiments with two standard judgmental forecasting datasets, with instances of our framework with two or three agents, empowered by six different base LLMs. We observe that combining evidence from agents can improve forecasting accuracy, especially in the case of three agents, while providing an explainable combination of evidence for claim verification.
Authors: Shengqi Xu, Xinpeng Zhou, Yabo Zhang, Ming Liu, Tao Liang, Tianyu Zhang, Yalong Bai, Zuxuan Wu, Wangmeng Zuo
Abstract: Personalized image preference assessment aims to evaluate an individual user's image preferences by relying only on a small set of reference images as prior information. Existing methods mainly focus on general preference assessment, training models with large-scale data to tackle well-defined tasks such as text-image alignment. However, these approaches struggle to handle personalized preference because user-specific data are scarce and not easily scalable, and individual tastes are often diverse and complex. To overcome these challenges, we introduce a common preference profile that serves as a bridge across users, allowing large-scale user data to be leveraged for training profile prediction and capturing complex personalized preferences. Building on this idea, we propose a reasoning-based personalized image preference assessment framework that follows a \textit{predict-then-assess} paradigm: it first predicts a user's preference profile from reference images, and then provides interpretable, multi-dimensional scores and assessments of candidate images based on the predicted profile. To support this, we first construct a large-scale Chain-of-Thought (CoT)-style personalized assessment dataset annotated with diverse user preference profiles and high-quality CoT-style reasoning, enabling explicit supervision of structured reasoning. Next, we adopt a two-stage training strategy: a cold-start supervised fine-tuning phase to empower the model with structured reasoning capabilities, followed by reinforcement learning to incentivize the model to explore more reasonable assessment paths and enhance generalization. Furthermore, we propose a similarity-aware prediction reward to encourage better prediction of the user's preference profile, which facilitates more reasonable assessments exploration. Extensive experiments demonstrate the superiority of the proposed method.
Authors: Haebin Seong, Sungmin Kim, Yongjun Cho, Myunchul Joe, Geunwoo Kim, Yubeen Park, Sunhoo Kim, Yoonshik Kim, Suhwan Choi, Jaeyoon Jung, Jiyong Youn, Jinmyung Kwak, Sunghee Ahn, Jaemin Lee, Younggil Do, Seungyeop Yi, Woojin Cheong, Minhyeok Oh, Minchan Kim, Seongjae Kang, Samwoo Seong, Youngjae Yu, Yunsung Lee
Abstract: While current navigation benchmarks prioritize task success in simplified settings, they neglect the multidimensional economic constraints essential for the real-world commercialization of autonomous delivery systems. We introduce CostNav, an Economic Navigation Benchmark that evaluates physical AI agents through comprehensive economic cost-revenue analysis aligned with real-world business operations. By integrating industry-standard data - such as SEC filings and AIS injury reports - with Isaac Sim's detailed collision and cargo dynamics, CostNav transcends simple task completion to accurately evaluate business value in complex, real-world scenarios. To our knowledge, CostNav is the first work to quantitatively expose the gap between navigation research metrics and commercial viability, revealing that optimizing for task success on a simplified task fundamentally differs from optimizing for real-world economic deployment. Our evaluation of rule-based Nav2 navigation shows that current approaches are not economically viable: the contribution margin is -22.81/run (AMCL) and -12.87/run (GPS), resulting in no break-even point. We challenge the community to develop navigation policies that achieve economic viability on CostNav. We remain method-agnostic, evaluating success solely on the metric of cost rather than the underlying architecture. All resources are available at https://github.com/worv-ai/CostNav.
Authors: Austin Spizzirri
Abstract: I argue that content-based AI value alignment--any approach that treats alignment as optimizing toward a formal value-object (reward function, utility function, constitutional principles, or learned preference representation)--cannot, by itself, produce robust alignment under capability scaling, distributional shift, and increasing autonomy. This limitation arises from three philosophical results: Hume's is-ought gap (behavioral data cannot entail normative conclusions), Berlin's value pluralism (human values are irreducibly plural and incommensurable), and the extended frame problem (any value encoding will misfit future contexts that advanced AI creates). I show that RLHF, Constitutional AI, inverse reinforcement learning, and cooperative assistance games each instantiate this specification trap, and that their failure modes are structural, not engineering limitations. Proposed escape routes--continual updating, meta-preferences, moral realism--relocate the trap rather than exit it. Drawing on Fischer and Ravizza's compatibilist theory, I argue that behavioral compliance does not constitute alignment: there is a principled distinction between simulated value-following and genuine reasons-responsiveness, and specification-based methods cannot produce the latter. The specification trap establishes a ceiling on content-based approaches, not their uselessness--but this ceiling becomes safety-critical at the capability frontier. The alignment problem must be reframed from value specification to value emergence.
Authors: Shengji Tang, Weihao Lin, Peng Ye, Jingqi Ye, Hao Li, Yiqun Zhang, Xiaosong Wang, Bo Zhang, Shuyue Hu, Tao Chen, Lei Bai, Wanli Ouyang
Abstract: Large Language Models (LLMs) have rapidly advanced, with Gemini-3-Pro setting a new performance milestone. In this work, we explore collective intelligence as an alternative to monolithic scaling, and demonstrate that open-source LLMs' collaboration can surpass Gemini-3-Pro. We first revisit LLM routing and aggregation at scale and identify three key bottlenecks: (1) current train-free routers are limited by a query-based paradigm focusing solely on textual similarity; (2) recent aggregation methods remain largely static, failing to select appropriate aggregators for different tasks;(3) the complementarity of routing and aggregation remains underutilized. To address these problems, we introduce JiSi, a novel framework designed to release the full potential of LLMs' collaboration through three innovations: (1) Query-Response Mixed Routing capturing both semantic information and problem difficulty; (2) Support-Set-based Aggregator Selection jointly evaluating the aggregation and domain capacity of aggregators; (3) Adaptive Routing-Aggregation Switch dynamically leveraging the advantages of routing and aggregation. Comprehensive experiments on nine benchmarks demonstrate that JiSi can surpass Gemini-3-Pro with only 47% costs by orchestrating ten open-source LLMs, while outperforming mainstream baselines. It suggests that collective intelligence represents a novel path towards Artificial General Intelligence (AGI).
Authors: Marc Lanctot, Kate Larson, Ian Gemp, Michael Kaisers
Abstract: As intelligent agents become more generally-capable, i.e. able to master a wide variety of tasks, the complexity and cost of properly evaluating them rises significantly. Tasks that assess specific capabilities of the agents can be correlated and stochastic, requiring many samples for accurate comparisons, leading to added costs. In this paper, we propose a formal definition and a conceptual framework for active evaluation of agents across multiple tasks, which assesses the performance of ranking algorithms as a function of number of evaluation data samples. Rather than curating, filtering, or compressing existing data sets as a preprocessing step, we propose an online framing: on every iteration, the ranking algorithm chooses the task and agents to sample scores from. Then, evaluation algorithms report a ranking of agents on each iteration and their performance is assessed with respect to the ground truth ranking over time. Several baselines are compared under different experimental contexts, with synthetic generated data and simulated online access to real evaluation data from Atari game-playing agents. We find that the classical Elo rating system -- while it suffers from well-known failure modes, in theory -- is a consistently reliable choice for efficient reduction of ranking error in practice. A recently-proposed method, Soft Condorcet Optimization, shows comparable performance to Elo on synthetic data and significantly outperforms Elo on real Atari agent evaluation. When task variation from the ground truth is high, selecting tasks based on proportional representation leads to higher rate of ranking error reduction.
Authors: Haoran Ye, Xuning He, Vincent Arak, Haonan Dong, Guojie Song
Abstract: The operational efficacy of large language models relies heavily on their inference-time context. This has established Context Engineering (CE) as a formal discipline for optimizing these inputs. Current CE methods rely on manually crafted harnesses, such as rigid generation-reflection workflows and predefined context schemas. They impose structural biases and restrict context optimization to a narrow, intuition-bound design space. To address this, we introduce Meta Context Engineering (MCE), a bi-level framework that supersedes static CE heuristics by co-evolving CE skills and context artifacts. In MCE iterations, a meta-level agent refines engineering skills via agentic crossover, a deliberative search over the history of skills, their executions, and evaluations. A base-level agent executes these skills, learns from training rollouts, and optimizes context as flexible files and code. We evaluate MCE across five disparate domains under offline and online settings. MCE demonstrates consistent performance gains, achieving 5.6--53.8% relative improvement over state-of-the-art agentic CE methods (mean of 16.9%), while maintaining superior context adaptability, transferability, and efficiency in both context usage and training.
Authors: Lakshya Gupta, Litao Li, Yizhe Liu, Sriram Ganapathi Subramanian, Kaheer Suleman, Zichen Zhang, Haoye Lu, Sumit Pasupalak
Abstract: Frontier large language models (LLMs) excel as autonomous agents in many domains, yet they remain untested in complex enterprise systems where hidden workflows create cascading effects across interconnected databases. Existing enterprise benchmarks evaluate surface-level agentic task completion similar to general consumer benchmarks, ignoring true challenges in enterprises, such as limited observability, large database state, and hidden workflows with cascading side effects. We introduce World of Workflows (WoW), a realistic ServiceNow-based environment incorporating 4,000+ business rules and 55 active workflows embedded in the system, alongside WoW-bench, a benchmark of 234 tasks evaluating constrained agentic task completion and enterprise dynamics modeling capabilities. We reveal two major takeaways: (1) Frontier LLMs suffer from dynamics blindness, consistently failing to predict the invisible, cascading side effects of their actions, which leads to silent constraint violations, and (2) reliability in opaque systems requires grounded world modeling, where agents must mentally simulate hidden state transitions to bridge the observability gap when high-fidelity feedback is unavailable. For reliable and useful enterprise agents, WoW motivates a new paradigm to explicitly learn system dynamics. We release our GitHub for setting up and evaluating WoW.
Authors: Bowen Cao, Dongdong Zhang, Yixia Li, Junpeng Liu, Shijue Huang, Chufan Shi, Hongyuan Lu, Yaokang Wu, Guanhua Chen, Wai Lam, Furu Wei
Abstract: Large language models now solve many benchmark math problems at near-expert levels, yet this progress has not fully translated into reliable performance in real-world applications. We study this gap through contextual mathematical reasoning, where the mathematical core must be formulated from descriptive scenarios. We introduce ContextMATH, a benchmark that repurposes AIME and MATH-500 problems into two contextual settings: Scenario Grounding (SG), which embeds abstract problems into realistic narratives without increasing reasoning complexity, and Complexity Scaling (CS), which transforms explicit conditions into sub-problems to capture how constraints often appear in practice. Evaluating 61 proprietary and open-source models, we observe sharp drops: on average, open-source models decline by 13 and 34 points on SG and CS, while proprietary models drop by 13 and 20. Error analysis shows that errors are dominated by incorrect problem formulation, with formulation accuracy declining as original problem difficulty increases. Correct formulation emerges as a prerequisite for success, and its sufficiency improves with model scale, indicating that larger models advance in both understanding and reasoning. Nevertheless, formulation and reasoning remain two complementary bottlenecks that limit contextual mathematical problem solving. Finally, we find that fine-tuning with scenario data improves performance, whereas formulation-only training is ineffective. However, performance gaps are only partially alleviated, highlighting contextual mathematical reasoning as a central unsolved challenge for LLMs.
Authors: Mengxuan Wang, Yuxin Chen, Gang Xu, Tao He, Hongjie Jiang, Ming Li
Abstract: Vision language models (VLMs) extend the reasoning capabilities of large language models (LLMs) to cross-modal settings, yet remain highly vulnerable to multimodal jailbreak attacks. Existing defenses predominantly rely on safety fine-tuning or aggressive token manipulations, incurring substantial training costs or significantly degrading utility. Recent research shows that LLMs inherently recognize unsafe content in text, and the incorporation of visual inputs in VLMs frequently dilutes risk-related signals. Motivated by this, we propose Risk Awareness Injection (RAI), a lightweight and training-free framework for safety calibration that restores LLM-like risk recognition by amplifying unsafe signals in VLMs. Specifically, RAI constructs an Unsafe Prototype Subspace from language embeddings and performs targeted modulation on selected high-risk visual tokens, explicitly activating safety-critical signals within the cross-modal feature space. This modulation restores the model's LLM-like ability to detect unsafe content from visual inputs, while preserving the semantic integrity of original tokens for cross-modal reasoning. Extensive experiments across multiple jailbreak and utility benchmarks demonstrate that RAI substantially reduces attack success rate without compromising task performance.
Authors: Guozhi Liu, Weiwei Lin, Tiansheng Huang, Ruichao Mo, Qi Mu, Xiumin Wang, Li Shen
Abstract: Harmful fine-tuning can invalidate safety alignment of large language models, exposing significant safety risks. In this paper, we utilize the attention sink mechanism to mitigate harmful fine-tuning. Specifically, we first measure a statistic named \emph{sink divergence} for each attention head and observe that \emph{different attention heads exhibit two different signs of sink divergence}. To understand its safety implications, we conduct experiments and find that the number of attention heads of positive sink divergence increases along with the increase of the model's harmfulness when undergoing harmful fine-tuning. Based on this finding, we propose a separable sink divergence hypothesis -- \emph{attention heads associating with learning harmful patterns during fine-tuning are separable by their sign of sink divergence}. Based on the hypothesis, we propose a fine-tuning-stage defense, dubbed Surgery. Surgery utilizes a regularizer for sink divergence suppression, which steers attention heads toward the negative sink divergence group, thereby reducing the model's tendency to learn and amplify harmful patterns. Extensive experiments demonstrate that Surgery improves defense performance by 5.90\%, 11.25\%, and 9.55\% on the BeaverTails, HarmBench, and SorryBench benchmarks, respectively. Source code is available on https://github.com/Lslland/Surgery.
Authors: Chris Zhu, Sasha Cui, Will Sanok Dufallo, Runzhi Jin, Zhen Xu, Linjun Zhang, Daylian Cain
Abstract: We present an in-depth evaluation of LLMs' ability to negotiate, a central business task that requires strategic reasoning, theory of mind, and economic value creation. To do so, we introduce PieArena, a large-scale negotiation benchmark grounded in multi-agent interactions over realistic scenarios drawn from an MBA negotiation course at an elite business school. We develop a statistically grounded ranking model for continuous negotiation payoffs that produces leaderboards with principled confidence intervals and corrects for experimental asymmetries. We find systematic evidence of human-expert-level performance in which a representative frontier language agent (GPT-5) matches or outperforms trained business-school students, despite a semester of general negotiation instruction and targeted coaching immediately prior to the task. We further study the effects of joint-intentionality agentic scaffolding and observe asymmetric gains, with large improvements for mid- and lower-tier LMs and diminishing returns for frontier LMs. Beyond deal outcomes, PieArena provides a multi-dimensional negotiation behavioral profile, revealing novel cross-model heterogeneity, masked by deal-outcome-only benchmarks, in deception, computation accuracy, instruction compliance, and perceived reputation. Overall, our results suggest that frontier language agents are already intellectually and psychologically capable of deployment in high-stakes economic settings, but deficiencies in robustness and trustworthiness remain open challenges.
Authors: Finn Rietz, Mart Karta\v{s}ev, Petter \"Ogren, Johannes A. Stork
Abstract: Behavior Trees (BTs) provide a structured and reactive framework for decision-making, commonly used to switch between sub-controllers based on environmental conditions. Reinforcement Learning (RL), on the other hand, can learn near-optimal controllers but sometimes struggles with sparse rewards, safe exploration, and long-horizon credit assignment. Combining BTs with RL has the potential for mutual benefit: a BT design encodes structured domain knowledge that can simplify RL training, while RL enables automatic learning of the controllers within BTs. However, naive integration of BTs and RL can lead to some controllers counteracting other controllers, possibly undoing previously achieved subgoals, thereby degrading the overall performance. To address this, we propose progress constraints, a novel mechanism where feasibility estimators constrain the allowed action set based on theoretical BT convergence results. Empirical evaluations in a 2D proof-of-concept and a high-fidelity warehouse environment demonstrate improved performance, sample efficiency, and constraint satisfaction, compared to prior methods of BT-RL integration.
Authors: Congcong Hu, Yuang Shi, Fan Huang, Yang Xiang, Zhou Ye, Ming Jin, Shiyu Wang
Abstract: Demand forecasting is a cornerstone of e-commerce operations, directly impacting inventory planning and fulfillment scheduling. However, existing forecasting systems often fail during high-impact periods such as flash sales, holiday campaigns, and sudden policy interventions, where demand patterns shift abruptly and unpredictably. In this paper, we introduce EventCast, a modular forecasting framework that integrates future event knowledge into time-series prediction. Unlike prior approaches that ignore future interventions or directly use large language models (LLMs) for numerical forecasting, EventCast leverages LLMs solely for event-driven reasoning. Unstructured business data, which covers campaigns, holiday schedules, and seller incentives, from existing operational databases, is processed by an LLM that converts it into interpretable textual summaries leveraging world knowledge for cultural nuances and novel event combinations. These summaries are fused with historical demand features within a dual-tower architecture, enabling accurate, explainable, and scalable forecasts. Deployed on real-world e-commerce scenarios spanning 4 countries of 160 regions over 10 months, EventCast achieves up to 86.9% and 97.7% improvement on MAE and MSE compared to the variant without event knowledge, and reduces MAE by up to 57.0% and MSE by 83.3% versus the best industrial baseline during event-driven periods. EventCast has deployed into real-world industrial pipelines since March 2025, offering a practical solution for improving operational decision-making in dynamic e-commerce environments.
Authors: Guanglong Sun, Hongwei Yan, Liyuan Wang, Zhiqi Kang, Shuang Cui, Hang Su, Jun Zhu, Yi Zhong
Abstract: To cope with uncertain changes of the external world, intelligent systems must continually learn from complex, evolving environments and respond in real time. This ability, collectively known as general continual learning (GCL), encapsulates practical challenges such as online datastreams and blurry task boundaries. Although leveraging pretrained models (PTMs) has greatly advanced conventional continual learning (CL), these methods remain limited in reconciling the diverse and temporally mixed information along a single pass, resulting in sub-optimal GCL performance. Inspired by meta-plasticity and reconstructive memory in neuroscience, we introduce here an innovative approach named Meta Post-Refinement (MePo) for PTMs-based GCL. This approach constructs pseudo task sequences from pretraining data and develops a bi-level meta-learning paradigm to refine the pretrained backbone, which serves as a prolonged pretraining phase but greatly facilitates rapid adaptation of representation learning to downstream GCL tasks. MePo further initializes a meta covariance matrix as the reference geometry of pretrained representation space, enabling GCL to exploit second-order statistics for robust output alignment. MePo serves as a plug-in strategy that achieves significant performance gains across a variety of GCL benchmarks and pretrained checkpoints in a rehearsal-free manner (e.g., 15.10\%, 13.36\%, and 12.56\% on CIFAR-100, ImageNet-R, and CUB-200 under Sup-21/1K). Our source code is available at \href{https://github.com/SunGL001/MePo}{MePo}
Authors: Yuhang Wang, Feiming Xu, Zheng Lin, Guangyu He, Yuzhe Huang, Haichang Gao, Zhenxing Niu, Shiguo Lian, Zhaoxiang Liu
Abstract: Although large language model (LLM)-based agents, exemplified by OpenClaw, are increasingly evolving from task-oriented systems into personalized AI assistants for solving complex real-world tasks, their practical deployment also introduces severe security risks. However, existing agent security research and evaluation frameworks primarily focus on synthetic or task-centric settings, and thus fail to accurately capture the attack surface and risk propagation mechanisms of personalized agents in real-world deployments. To address this gap, we propose Personalized Agent Security Bench (PASB), an end-to-end security evaluation framework tailored for real-world personalized agents. Building upon existing agent attack paradigms, PASB incorporates personalized usage scenarios, realistic toolchains, and long-horizon interactions, enabling black-box, end-to-end security evaluation on real systems. Using OpenClaw as a representative case study, we systematically evaluate its security across multiple personalized scenarios, tool capabilities, and attack types. Our results indicate that OpenClaw exhibits critical vulnerabilities at different execution stages, including user prompt processing, tool usage, and memory retrieval, highlighting substantial security risks in personalized agent deployments. The code for the proposed PASB framework is available at https://github.com/AstorYH/PASB.
Authors: Xinhai Sun
Abstract: Modern large language models (LLMs) are often evaluated and deployed under a one-shot, greedy inference protocol, especially in professional settings that require deterministic behavior. This regime can systematically under-estimate a fixed model's true capability: many errors arise not from missing knowledge, but from premature commitment under internal ambiguity. We introduce Reinforcement Inference, an entropy-aware inference-time control strategy that uses the model's own uncertainty to selectively invoke a second, more deliberate reasoning attempt, enabling stronger performance without any retraining. On 12,032 MMLU-Pro questions across 14 subjects, using DeepSeek-v3.2 with deterministic decoding in a zero-shot setting, Reinforcement Inference improves accuracy from 60.72% to 84.03%, while only incurring 61.06% additional inference calls. A 100% re-asking ablation reaches 84.35%, indicating that uncertainty-aware selection captures most of the attainable improvement with substantially less compute. Moreover, a prompt-only ablation underperforms the baseline, suggesting that the gains are not explained by generic prompting alone. Beyond providing a practical inference-time upgrade, our results suggest a broader entropy-aware paradigm for measuring and expanding model capability: because modern decoder-based models generate outputs autoregressively, entropy and related confidence measures arise naturally as first-class control signals during generation. The resulting gap between one-pass greedy inference and uncertainty-conditioned deliberation offers a diagnostic lens on an LLM's latent reasoning horizon and motivates future training objectives that explicitly constrain correctness--confidence alignment.
Authors: Aditya Gulati, Nuria Oliver
Abstract: As chatbots increasingly blur the boundary between automated systems and human conversation, the foundations of trust in these systems warrant closer examination. While regulatory and policy frameworks tend to define trust in normative terms, the trust users place in chatbots often emerges from behavioral mechanisms. In many cases, this trust is not earned through demonstrated trustworthiness but is instead shaped by interactional design choices that leverage cognitive biases to influence user behavior. Based on this observation, we propose reframing chatbots not as companions or assistants, but as highly skilled salespeople whose objectives are determined by the deploying organization. We argue that the coexistence of competing notions of "trust" under a shared term obscures important distinctions between psychological trust formation and normative trustworthiness. Addressing this gap requires further research and stronger support mechanisms to help users appropriately calibrate trust in conversational AI systems.
Authors: Andr\'es Holgado-S\'anchez, Peter Vamplew, Richard Dazeley, Sascha Ossowski, Holger Billhardt
Abstract: Value-aware AI should recognise human values and adapt to the value systems (value-based preferences) of different users. This requires operationalization of values, which can be prone to misspecification. The social nature of values demands their representation to adhere to multiple users while value systems are diverse, yet exhibit patterns among groups. In sequential decision making, efforts have been made towards personalization for different goals or values from demonstrations of diverse agents. However, these approaches demand manually designed features or lack value-based interpretability and/or adaptability to diverse user preferences. We propose algorithms for learning models of value alignment and value systems for a society of agents in Markov Decision Processes (MDPs), based on clustering and preference-based multi-objective reinforcement learning (PbMORL). We jointly learn socially-derived value alignment models (groundings) and a set of value systems that concisely represent different groups of users (clusters) in a society. Each cluster consists of a value system representing the value-based preferences of its members and an approximately Pareto-optimal policy that reflects behaviours aligned with this value system. We evaluate our method against a state-of-the-art PbMORL algorithm and baselines on two MDPs with human values.
Authors: Furong Jia, Ling Dai, Wenjin Deng, Fan Zhang, Chen Hu, Daxin Jiang, Yu Liu
Abstract: Large Vision-Language Models (LVLMs) have demonstrated strong reasoning capabilities in geo-localization, yet they often struggle in real-world scenarios where visual cues are sparse, long-tailed, and highly ambiguous. Previous approaches, bound by internal knowledge, often fail to provide verifiable results, yielding confident but ungrounded predictions when faced with confounded evidence. To address these challenges, we propose SpotAgent, a framework that formalizes geo-localization into an agentic reasoning process that leverages expert-level reasoning to synergize visual interpretation with tool-assisted verification. SpotAgent actively explores and verifies visual cues by leveraging external tools (e.g., web search, maps) through a ReAct diagram. We introduce a 3-stage post-training pipeline starting with a Supervised Fine-Tuning (SFT) stage for basic alignment, followed by an Agentic Cold Start phase utilizing high-quality trajectories synthesized via a Multi-Agent framework, aiming to instill tool-calling expertise. Subsequently, the model's reasoning capabilities are refined through Reinforcement Learning. We propose a Spatially-Aware Dynamic Filtering strategy to enhance the efficiency of the RL stage by prioritizing learnable samples based on spatial difficulty. Extensive experiments on standard benchmarks demonstrate that SpotAgent achieves state-of-the-art performance, effectively mitigating hallucinations while delivering precise and verifiable geo-localization.
Authors: Shiwei Lyu, Xidong Wang, Lei Liu, Hao Zhu, Chaohe Zhang, Jian Wang, Jinjie Gu, Benyou Wang, Yue Shen
Abstract: Although large language models (LLMs) demonstrate expert-level medical knowledge, aligning their open-ended outputs with fine-grained clinician preferences remains challenging. Existing methods often rely on coarse objectives or unreliable automated judges that are weakly grounded in professional guidelines. We propose a two-stage framework to address this gap. First, we introduce HealthRubrics, a dataset of 7,034 physician-verified preference examples in which clinicians refine LLM-drafted rubrics to meet rigorous medical standards. Second, we distill these rubrics into HealthPrinciples: 119 broadly reusable, clinically grounded principles organized by clinical dimensions, enabling scalable supervision beyond manual annotation. We use HealthPrinciples for (1) offline alignment by synthesizing rubrics for unlabeled queries and (2) an inference-time tool for guided self-revision. A 30B-A3B model trained with our framework achieves 33.4% on HealthBench-Hard, outperforming much larger models including Deepseek-R1 and o3, establishing a resource-efficient baseline for clinical alignment.
Authors: Richard Bornemann, Pierluigi Vito Amadori, Antoine Cully
Abstract: Developing agents capable of open-endedly discovering and learning novel skills is a grand challenge in Artificial Intelligence. While reinforcement learning offers a powerful framework for training agents to master complex skills, it typically relies on hand-designed reward functions. This is infeasible for open-ended skill discovery, where the set of meaningful skills is not known a priori. While recent methods have shown promising results towards automating reward function design, they remain limited to refining rewards for pre-defined tasks. To address this limitation, we introduce Continuous Open-ended Discovery and Evolution of Skills as Hierarchical Reward Programs (CODE-SHARP), a novel framework leveraging Foundation Models (FM) to open-endedly expand and refine a hierarchical skill archive, structured as a directed graph of executable reward functions in code. We show that a goal-conditioned agent trained exclusively on the rewards generated by the discovered SHARP skills learns to solve increasingly long-horizon goals in the Craftax environment. When composed by a high-level FM-based planner, the discovered skills enable a single goal-conditioned agent to solve complex, long-horizon tasks, outperforming both pretrained agents and task-specific expert policies by over $134$% on average. We will open-source our code and provide additional videos at https://sites.google.com/view/code-sharp/homepage.
Authors: Zhaoyang Wang, Canwen Xu, Boyi Liu, Yite Wang, Siwei Han, Zhewei Yao, Huaxiu Yao, Yuxiong He
Abstract: Recent advances in large language model (LLM) have empowered autonomous agents to perform complex tasks that require multi-turn interactions with tools and environments. However, scaling such agent training is limited by the lack of diverse and reliable environments. In this paper, we propose Agent World Model (AWM), a fully synthetic environment generation pipeline. Using this pipeline, we scale to 1,000 environments covering everyday scenarios, in which agents can interact with rich toolsets (35 tools per environment on average) and obtain high-quality observations. Notably, these environments are code-driven and backed by databases, providing more reliable and consistent state transitions than environments simulated by LLMs. Moreover, they enable more efficient agent interaction compared with collecting trajectories from realistic environments. To demonstrate the effectiveness of this resource, we perform large-scale reinforcement learning for multi-turn tool-use agents. Thanks to the fully executable environments and accessible database states, we can also design reliable reward functions. Experiments on three benchmarks show that training exclusively in synthetic environments, rather than benchmark-specific ones, yields strong out-of-distribution generalization. The code is available at https://github.com/Snowflake-Labs/agent-world-model.
Authors: Daniel Fern\'andez-Gonz\'alez
Abstract: Structured sentiment analysis (SSA) aims to automatically extract people's opinions from a text in natural language and adequately represent that information in a graph structure. One of the most accurate methods for performing SSA was recently proposed and consists of approaching it as a dependency graph parsing task. Although we can find in the literature how transition-based algorithms excel in different dependency graph parsing tasks in terms of accuracy and efficiency, all proposed attempts to tackle SSA following that approach were based on graph-based models. In this article, we present the first transition-based method to address SSA as dependency graph parsing. Specifically, we design a transition system that processes the input text in a left-to-right pass, incrementally generating the graph structure containing all identified opinions. To effectively implement our final transition-based model, we resort to a Pointer Network architecture as a backbone. From an extensive evaluation, we demonstrate that our model offers the best performance to date in practically all cases among prior dependency-based methods, and surpasses recent task-specific techniques on the most challenging datasets. We additionally include an in-depth analysis and empirically prove that the average-case time complexity of our approach is quadratic in the sentence length, being more efficient than top-performing graph-based parsers.
Authors: Yoav Kolumbus, Joe Halpern, \'Eva Tardos
Abstract: In repeated games, such as auctions, players rely on autonomous learning agents to choose their actions. We study settings in which players have their agents make monetary transfers to other agents during play at their own expense, in order to influence learning dynamics in their favor. Our goal is to understand when players have incentives to use such payments, how payments between agents affect learning outcomes, and what the resulting implications are for welfare and its distribution. We propose a simple game-theoretic model to capture the incentive structure of such scenarios. We find that, quite generally, abstaining from payments is not robust to strategic deviations by users of learning agents: self-interested players benefit from having their agents make payments to other learners. In a broad class of games, such endogenous payments between learning agents lead to higher welfare for all players. In first- and second-price auctions, equilibria of the induced "payment-policy game" lead to highly collusive learning outcomes, with low or vanishing revenue for the auctioneer. These results highlight a fundamental challenge for mechanism design, as well as for regulatory policies, in environments where learning agents may interact in the digital ecosystem beyond a mechanism's boundaries.
Authors: Wilson G. Gregory, Josu\'e Tonelli-Cueto, Nicholas F. Marshall, Andrew S. Lee, Soledad Villar
Abstract: Tensors are a fundamental data structure for many scientific contexts, such as time series analysis, materials science, and physics, among many others. Improving our ability to produce and handle tensors is essential to efficiently address problems in these domains. In this paper, we show how to exploit the underlying symmetries of functions that map tensors to tensors. More concretely, we develop universally expressive equivariant machine learning architectures on tensors that exploit that, in many cases, these tensor functions are equivariant with respect to the diagonal action of the orthogonal, Lorentz, and/or symplectic groups. We showcase our results on three problems coming from material science, theoretical computer science, and time series analysis. For time series, we combine our method with the increasingly popular path signatures approach, which is also invariant with respect to reparameterizations. Our numerical experiments show that our equivariant models perform better than corresponding non-equivariant baselines.
Authors: Jiayi Lin, Yanlin Wang, Yibiao Yang, Lei Zhang, Yutao Xie
Abstract: Recent advances in large-scale code generation models have led to remarkable progress in producing high-quality code. These models are trained in a self-supervised manner on extensive unlabeled code corpora using a decoder-only architecture. However, despite their generative strength, decoder-only models often exhibit limited performance on code understanding tasks such as code search and clone detection, primarily due to their generation-oriented training objectives. While training large encoder-only models from scratch on massive code datasets can improve understanding ability but remains computationally expensive and time-consuming. In this paper, we explore a more efficient alternative by transferring knowledge from pre-trained decoder-only code generation models to code understanding tasks. We investigate how decoder-only architectures can be effectively adapted to learn discriminative and semantically meaningful code representations. To this end, we propose CL4D, a contrastive learning framework tailored to strengthen the representation capabilities of decoder-only models. Extensive experiments on multiple benchmark datasets demonstrate that CL4D achieves competitive or superior performance compared to existing methods on representative code understanding tasks, including code search and clone detection. Further analysis reveals that CL4D substantially improves the semantic alignment of code representations by reducing the distance between semantically similar code snippets. These findings highlight the feasibility of leveraging decoder-only models as a unified backbone for both code generation and understanding.
Authors: Jialiang Wang, Hanmo Liu, Shimin Di, Zhili Wang, Jiachuan Wang, Lei Chen, Xiaofang Zhou
Abstract: High-level automation is increasingly critical in AI, driven by rapid advances in large language models (LLMs) and AI agents. However, LLMs, despite their general reasoning power, struggle significantly in specialized, data-sensitive tasks such as designing Graph Neural Networks (GNNs). This difficulty arises from (1) the inherent knowledge gaps in modeling the intricate, varying relationships between graph properties and suitable architectures and (2) the external noise from misleading descriptive inputs, often resulting in generic or even misleading model suggestions. Achieving proficiency in designing data-aware models -- defined as the meta-level capability to systematically accumulate, interpret, and apply data-specific design knowledge -- remains challenging for existing automated approaches, due to their inefficient construction and application of meta-knowledge. To achieve meta-level proficiency, we propose DesiGNN, a knowledge-centered framework that systematically converts past model design experience into structured, fine-grained knowledge priors well-suited for meta-learning with LLMs. To account for the inherent variability and external noise, DesiGNN aligns empirical property filtering from extensive benchmarks with adaptive elicitation of literature insights via LLMs. By constructing a solid meta-knowledge between unseen graph understanding and known effective architecture patterns, DesiGNN can deliver top-5.77% initial model proposals for unseen datasets within seconds and achieve consistently superior performance with minimal search cost compared to baselines.
Authors: Elia Bonetto, Aamir Ahmad
Abstract: Collecting and labeling large real-world wild animal datasets is impractical, costly, error-prone, and labor-intensive. For animal monitoring tasks, as detection, tracking, and pose estimation, out-of-distribution viewpoints (e.g. aerial) are also typically needed but rarely found in publicly available datasets. To solve this, existing approaches synthesize data with simplistic techniques that then necessitate strategies to bridge the synthetic-to-real gap. Therefore, real images, style constraints, complex animal models, or pre-trained networks are often leveraged. In contrast, we generate a fully synthetic dataset using a 3D photorealistic simulator and demonstrate that it can eliminate such needs for detecting and estimating 2D poses of wild zebras. Moreover, existing top-down 2D pose estimation approaches using synthetic data assume reliable detection models. However, these often fail in out-of-distribution scenarios, e.g. those that include wildlife or aerial imagery. Our method overcomes this by enabling the training of both tasks using the same synthetic dataset. Through extensive benchmarks, we show that models trained from scratch exclusively on our synthetic data generalize well to real images. We perform these using multiple real-world and synthetic datasets, pre-trained and randomly initialized backbones, and different image resolutions. Code, results, models, and data can be found athttps://zebrapose.is.tue.mpg.de/.
Authors: Simon Sinong Zhan, Philip Wang, Qingyuan Wu, Yixuan Wang, Ruochen Jiao, Chao Huang, Qi Zhu
Abstract: In this paper, we aim to tackle the limitation of the Adversarial Inverse Reinforcement Learning (AIRL) method in stochastic environments where theoretical results cannot hold and performance is degraded. To address this issue, we propose a novel method which infuses the dynamics information into the reward shaping with the theoretical guarantee for the induced optimal policy in the stochastic environments. Incorporating our novel model-enhanced rewards, we present a novel Model-Enhanced AIRL framework, which integrates transition model estimation directly into reward shaping. Furthermore, we provide a comprehensive theoretical analysis of the reward error bound and performance difference bound for our method. The experimental results in MuJoCo benchmarks show that our method can achieve superior performance in stochastic environments and competitive performance in deterministic environments, with significant improvement in sample efficiency, compared to existing baselines.
Authors: Jiankun Wei, Abdulrahman Abdulrazzag, Tianchen Zhang, Adel Muursepp, Gururaj Saileshwar
Abstract: Deployed large language models (LLMs) often rely on speculative decoding, a technique that generates and verifies multiple candidate tokens in parallel, to improve throughput and latency. In this work, we reveal a new side-channel whereby input-dependent patterns of correct and incorrect speculations can be inferred by monitoring per-iteration token counts or packet sizes. In evaluations using research prototypes and production-grade vLLM serving frameworks, we show that an adversary monitoring these patterns can fingerprint user queries (from a set of 50 prompts) with over 75% accuracy across four speculative-decoding schemes at temperature 0.3: REST (100%), LADE (91.6%), BiLD (95.2%), and EAGLE (77.6%). Even at temperature 1.0, accuracy remains far above the 2% random baseline - REST (99.6%), LADE (61.2%), BiLD (63.6%), and EAGLE (24%). We also show the capability of the attacker to leak confidential datastore contents used for prediction at rates exceeding 25 tokens/sec. To defend against these, we propose and evaluate a suite of mitigations, including packet padding and iteration-wise token aggregation.
Authors: Hao Shu
Abstract: Edge detection (ED) is a fundamental perceptual process in computer vision, forming the structural basis for high-level reasoning tasks such as segmentation, recognition, and scene understanding. Despite substantial progress achieved by deep neural networks, most ED models attain high numerical accuracy but fail to produce visually sharp and perceptually consistent edges, thereby limiting their reliability in intelligent vision systems. To address this issue, this study introduces the Symmetrization Weighted Binary Cross-Entropy (SWBCE) loss, a perception-inspired formulation that extends the conventional WBCE by incorporating prediction-guided symmetry. SWBCE explicitly models the perceptual asymmetry in human edge recognition, wherein edge decisions require stronger evidence than non-edge ones, aligning the optimization process with human perceptual discrimination. The resulting symmetric learning mechanism jointly enhances edge recall and suppresses false positives, achieving a superior balance between quantitative accuracy and perceptual fidelity. Extensive experiments across multiple benchmark datasets and representative ED architectures demonstrate that SWBCE can outperform existing loss functions in both numerical evaluation and visual quality. Particularly with the HED-EES model, the SSIM can be improved by about 15% on BRIND, and in all experiments, training by SWBCE consistently obtains the best perceptual results. Beyond edge detection, the proposed perceptual loss offers a generalizable optimization principle for soft computing and neural learning systems, particularly in scenarios where asymmetric perceptual reasoning plays a critical role.
Authors: Akshay Kudva, Wei-Ting Tang, Joel A. Paulson
Abstract: Designing modern industrial systems requires balancing several competing objectives, such as profitability, resilience, and sustainability, while accounting for complex interactions between technological, economic, and environmental factors. Multi-objective optimization (MOO) methods are commonly used to navigate these tradeoffs, but selecting the appropriate algorithm to tackle these problems is often unclear, particularly when system representations vary from fully equation-based (white-box) to entirely data-driven (black-box) models. While grey-box MOO methods attempt to bridge this gap, they typically impose rigid assumptions on system structure, requiring models to conform to the underlying structural assumptions of the solver rather than the solver adapting to the natural representation of the system of interest. In this chapter, we introduce a unifying approach to grey-box MOO by leveraging network representations, which provide a general and flexible framework for modeling interconnected systems as a series of function nodes that share various inputs and outputs. Specifically, we propose MOBONS, a novel Bayesian optimization-inspired algorithm that can efficiently optimize general function networks, including those with cyclic dependencies, enabling the modeling of feedback loops, recycle streams, and multi-scale simulations - features that existing methods fail to capture. Furthermore, MOBONS incorporates constraints, supports parallel evaluations, and preserves the sample efficiency of Bayesian optimization while leveraging network structure for improved scalability. We demonstrate the effectiveness of MOBONS through two case studies, including one related to sustainable process design. By enabling efficient MOO under general graph representations, MOBONS has the potential to significantly enhance the design of more profitable, resilient, and sustainable engineering systems.
Authors: Yu Yan, Sheng Sun, Zenghao Duan, Teli Liu, Min Liu, Zhiyi Yin, Jingyu Lei, Qi Li
Abstract: Current studies have exposed the risk of Large Language Models (LLMs) generating harmful content by jailbreak attacks. However, they overlook that the direct generation of harmful content from scratch is more difficult than inducing LLM to calibrate benign content into harmful forms. In our study, we introduce a novel attack framework that exploits AdVersArial meTAphoR (AVATAR) to induce the LLM to calibrate malicious metaphors for jailbreaking. Specifically, to answer harmful queries, AVATAR adaptively identifies a set of benign but logically related metaphors as the initial seed. Then, driven by these metaphors, the target LLM is induced to reason and calibrate about the metaphorical content, thus jailbroken by either directly outputting harmful responses or calibrating residuals between metaphorical and professional harmful content. Experimental results demonstrate that AVATAR can effectively and transferable jailbreak LLMs and achieve a state-of-the-art attack success rate across multiple advanced LLMs.
Authors: Jiachun Li, Pengfei Cao, Zhuoran Jin, Yubo Chen, Jiexin Xu, Huaijun Li, Xiaojian Jiang, Kang Liu, Jun Zhao
Abstract: Inference-time scaling techniques have shown promise in enhancing the reasoning capabilities of large language models (LLMs). While recent research has primarily focused on training-time optimization, our work highlights inference-time reward model (RM)-based reasoning as a critical yet overlooked avenue. In this paper, we conduct a systematic analysis of RM behavior across downstream reasoning tasks, revealing three key limitations: (1) RM can impair performance on simple questions, (2) its discriminative ability declines with increased sampling, and (3) high search diversity undermines RM performance. To address these issues, we propose CRISP (Clustered Reward Integration with Stepwise Prefixing), a novel inference-time algorithm that clusters generated reasoning paths by final answers, aggregates reward signals at the cluster level, and adaptively updates prefix prompts to guide generation. Experimental results demonstrate that CRISP significantly enhances LLM reasoning performance, achieving up to 5% accuracy improvement over other RM-based inference methods and an average of 10% gain over advanced reasoning models.
Authors: Yan Wang, Shijie Zhao, Kexin Zhang, Junlin Li, Li Zhang
Abstract: Although recent research applying text-to-image (T2I) diffusion models to real-world super-resolution (SR) has achieved remarkable progress, the misalignment of their targets leads to a suboptimal trade-off between inference speed and detail fidelity. Specifically, the T2I task requires multiple inference steps to synthesize images matching to prompts and reduces the latent dimension to lower generating difficulty. Contrariwise, SR can restore high-frequency details in fewer inference steps, but it necessitates a more reliable variational auto-encoder (VAE) to preserve input information. However, most diffusion-based SRs are multistep and use 4-channel VAEs, while existing models with 16-channel VAEs are overqualified diffusion transformers, e.g., FLUX (12B). To align the target, we present a one-step diffusion model for generative detail restoration, GenDR, distilled from a tailored diffusion model with a larger latent space. In detail, we train a new SD2.1-VAE16 (0.9B) via representation alignment to expand the latent space without increasing the model size. Regarding step distillation, we propose consistent score identity distillation (CiD) that incorporates SR task-specific loss into score distillation to leverage more SR priors and align the training target. Furthermore, we extend CiD with adversarial learning and representation alignment (CiDA) to enhance perceptual quality and accelerate training. We also polish the pipeline to achieve a more efficient inference. Experimental results demonstrate that GenDR achieves state-of-the-art performance in both quantitative metrics and visual fidelity.
Authors: Dan Leonte, Aamal Hussain, Raphael Huser, Francesco Belardinelli, Dario Paccagnan
Abstract: Beyond specific settings, many multi-agent learning algorithms fail to converge to an equilibrium solution, instead displaying complex, non-stationary behaviours such as recurrent or chaotic orbits. In fact, recent literature suggests that such complex behaviours are likely to occur when the number of agents increases. In this paper, we study Q-learning dynamics in network polymatrix normal-form games where the network structure is drawn from classical random graph models. In particular, we focus on the Erd\H{o}s-R\'enyi model, which is used to analyze connectivity in distributed systems, and the Stochastic Block model, which generalizes the above by accounting for community structures that naturally arise in multi-agent systems. In each setting, we establish sufficient conditions under which the agents' joint strategies converge to a unique equilibrium. We investigate how this condition depends on the exploration rates, payoff matrices and, crucially, the probabilities of interaction between network agents. We validate our theoretical findings through numerical simulations and demonstrate that convergence can be reliably achieved in many-agent systems, provided interactions in the network are controlled.
Authors: Philipp D. Siedler, Ian Gemp
Abstract: In complex multi-agent environments, achieving efficient learning and desirable behaviours is a significant challenge for Multi-Agent Reinforcement Learning (MARL) systems. This work explores the potential of combining MARL with Large Language Model (LLM)-mediated interventions to guide agents toward more desirable behaviours. Specifically, we investigate how LLMs can be used to interpret and facilitate interventions that shape the learning trajectories of multiple agents. We experimented with two types of interventions, referred to as controllers: a Natural Language (NL) Controller and a Rule-Based (RB) Controller. The RB Controller showed a stronger impact than the NL Controller, which uses a small (7B/8B) LLM to simulate human-like interventions. Our findings indicate that agents particularly benefit from early interventions, leading to more efficient training and higher performance. Both intervention types outperform the baseline without interventions, highlighting the potential of LLM-mediated guidance to accelerate training and enhance MARL performance in challenging environments.
Authors: Jialin Chen, Aosong Feng, Ziyu Zhao, Juan Garza, Gaukhar Nurbek, Cheng Qin, Ali Maatouk, Leandros Tassiulas, Yifeng Gao, Rex Ying
Abstract: Understanding the relationship between textual news and time-series evolution is a critical yet under-explored challenge in applied data science. While multimodal learning has gained traction, existing multimodal time-series datasets fall short in evaluating cross-modal reasoning and complex question answering, which are essential for capturing complex interactions between narrative information and temporal patterns. To bridge this gap, we introduce Multimodal Time Series Benchmark (MTBench), a large-scale benchmark designed to evaluate large language models (LLMs) on time series and text understanding across financial and weather domains. MTbench comprises paired time series and textual data, including financial news with corresponding stock price movements and weather reports aligned with historical temperature records. Unlike existing benchmarks that focus on isolated modalities, MTbench provides a comprehensive testbed for models to jointly reason over structured numerical trends and unstructured textual narratives. The richness of MTbench enables formulation of diverse tasks that require a deep understanding of both text and time-series data, including time-series forecasting, semantic and technical trend analysis, and news-driven question answering (QA). These tasks target the model's ability to capture temporal dependencies, extract key insights from textual context, and integrate cross-modal information. We evaluate state-of-the-art LLMs on MTbench, analyzing their effectiveness in modeling the complex relationships between news narratives and temporal patterns. Our findings reveal significant challenges in current models, including difficulties in capturing long-term dependencies, interpreting causality in financial and weather trends, and effectively fusing multimodal information.
Authors: Amna Naeem, Jawad Ahmad, Muazzam A. Khan, Aizaz Ahmad Khattak, Muhammad Shahbaz Khan
Abstract: The ever-increasing security vulnerabilities in the Internet-of-Things (IoT) systems require improved threat detection approaches. This paper presents a compact and efficient approach to detect botnet attacks by employing an integrated approach that consists of traffic pattern analysis, temporal support learning, and focused feature extraction. The proposed attention-based model benefits from a hybrid CNN-BiLSTM architecture and achieves 99% classification accuracy in detecting botnet attacks utilizing the N-BaIoT dataset, while maintaining high precision and recall across various scenarios. The proposed model's performance is further validated by key parameters, such as Mathews Correlation Coefficient and Cohen's kappa Correlation Coefficient. The close-to-ideal results for these parameters demonstrate the proposed model's ability to detect botnet attacks accurately and efficiently in practical settings and on unseen data. The proposed model proved to be a powerful defense mechanism for IoT networks to face emerging security challenges.
Authors: Chaoqi Liu, Yunzhu Li, Kris Hauser
Abstract: Predictive models can be particularly helpful for robots to effectively manipulate terrains in construction sites and extraterrestrial surfaces. However, terrain state representations become extremely high-dimensional especially to capture fine-resolution details and when depth is unknown or unbounded. This paper introduces a learning-based approach for terrain dynamics modeling and manipulation, leveraging the Graph-based Neural Dynamics (GBND) framework to represent terrain deformation as motion of a graph of particles. Based on the principle that the moving portion of a terrain is usually localized, our approach builds a large terrain graph (potentially millions of particles) but only identifies a very small active subgraph (hundreds of particles) for predicting the outcomes of robot-terrain interaction. To minimize the size of the active subgraph we introduce a learning-based approach that identifies a small region of interest (RoI) based on the robot's control inputs and the current scene. We also introduce a novel domain boundary feature encoding that allows GBNDs to perform accurate dynamics prediction in the RoI interior while avoiding particle penetration through RoI boundaries. Our proposed method is both orders of magnitude faster than naive GBND and it achieves better overall prediction accuracy. We further evaluated our framework on excavation and shaping tasks on terrain with different granularity.
Authors: Xixuan Hao, Yutian Jiang, Xingchen Zou, Jiabo Liu, Yifang Yin, Song Gao, Flora Salim, Tianrui Li, Yuxuan Liang
Abstract: The ability to transform location-centric geospatial data into meaningful computational representations has become fundamental to modern spatial analysis and decision-making. Geospatial Representation Learning (GRL), the process of automatically extracting latent structures and semantic patterns from geographic data, is undergoing a profound transformation through two successive technological revolutions: the deep learning breakthrough and the emerging large language model (LLM) paradigm. While deep neural networks (DNNs) have demonstrated remarkable success in automated feature extraction from structured and semi-structured geospatial data (e.g., satellite imagery, GPS trajectories), the recent integration of LLMs introduces transformative capabilities for cross-modal geospatial reasoning and unstructured geo-textual data processing. This survey presents a comprehensive review of geospatial representation learning across both technological eras, organizing them into a structured taxonomy based on the complete pipeline comprising: (1) data perspective, (2) methodological perspective, and (3) application perspective. We also highlight current advancements, discuss existing limitations, and propose potential future research directions in the LLM and foundation model era. This work offers a thorough exploration of the field and provides a roadmap for further innovation in GRL. The summary of the up-to-date paper list can be found in https://github.com/CityMind-Lab/Awesome-Geospatial-Representation-Learning and will undergo continuous updates.
URLs: https://github.com/CityMind-Lab/Awesome-Geospatial-Representation-Learning
Authors: Peimian Du, Jiabin Liu, Xiaowei Jin, Wangmeng Zuo, Hui Li
Abstract: This research confronts the challenge of substantial physical equation discrepancies encountered in the generation of spatiotemporal physical fields through data-driven trained models. A spatiotemporal physical field generation model, named HMT-PF, is developed based on the hybrid Mamba-Transformer architecture, incorporating unstructured grid information as input. A fine-tuning block, enhanced with physical information, is introduced to effectively reduce the physical equation discrepancies. The physical equation residuals are computed through a point query mechanism for efficient gradient evaluation, then encoded into latent space for refinement. The fine-tuning process employs a self-supervised learning approach to achieve physical consistency while maintaining essential field characteristics. Results show that the hybrid Mamba-Transformer model achieves good performance in generating spatiotemporal fields, while the physics-informed fine-tuning mechanism further reduces significant physical errors effectively. A MSE-R evaluation method is developed to assess the accuracy and realism of physical field generation.
Authors: Feijiang Han, Xiaodong Yu, Jianheng Tang, Delip Rao, Weihua Du, Lyle Ungar
Abstract: Token-level attention tuning, a class of training-free methods including Post-hoc Attention Steering (PASTA) and Attention Calibration (ACT), has emerged as a promising approach for improving frozen LLMs via interpretable interventions. However, these methods rely on auxiliary heuristics to identify important task-specific tokens, which can introduce bias and limit applicability when token importance is ambiguous or when optimized kernels make attention maps inaccessible. We propose a simpler alternative: intervening only on the initial token (e.g., BOS in LLaMA). We theoretically show that adding lightweight biases to this token's attention logits systematically shifts and reshapes downstream attention patterns - an effect amplified by its natural role as an attention sink. Empirically, we find that this tuning can improve LLM performance and better elicit pretrained knowledge, with stronger effects in early layers and distinct scaling preferences across attention heads. Building on these findings, we introduce ZeroTuning, a training-free method that improves LLM performance by applying head-specific attention adjustments to the initial token, requiring no parameter updates. We present two variants: a supervised mode that calibrates on validation examples, and an unsupervised mode that directly minimizes output entropy. ZeroTuning requires no KV-cache or decoding changes and is kernel-agnostic (works with SDPA and FlashAttention). It requires only four lines of modification to the standard LlamaAttention code, achieves gains across 15 datasets, and outperforms prior, more complex methods. For example, on Llama-3.1-8B, it yields relative improvements of 19.9% on classification, 4.5% on question answering, and 2.1% on dialogue. ZeroTuning also works out of the box with quantized inference and maintains its improvements as context length increases.
Authors: Yu-Ting Lee, Fu-Chieh Chang, Yu-En Shu, Hui-Ying Shih, Pei-Yuan Wu
Abstract: Intrinsic self-correction refers to the phenomenon where a language model refines its own outputs purely through prompting, without external feedback or parameter updates. While this approach improves performance across diverse tasks, its mechanism remains unclear. We show that intrinsic self-correction functions by steering hidden representations along interpretable latent directions, as evidenced by both alignment analysis and activation interventions. To achieve this, we analyze intrinsic self-correction via the representation shift induced by prompting. In parallel, we construct interpretable latent directions with contrastive pairs and verify the causal effect of these directions via activation addition. Evaluating six open-source LLMs, our results demonstrate that prompt-induced representation shifts in text detoxification and text toxification consistently align with latent directions constructed from contrastive pairs. In detoxification, the shifts align with the non-toxic direction; in toxification, they align with the toxic direction. These findings suggest that representation steering is the mechanistic driver of intrinsic self-correction. Our analysis highlights that understanding model internals offers a direct route to analyzing the mechanisms of prompt-driven LLM behaviors.
Authors: Ruizhe Li, Chen Chen, Yuchen Hu, Yanjun Gao, Xi Wang, Emine Yilmaz
Abstract: Retrieval-Augmented Generation (RAG) leverages large language models (LLMs) combined with external contexts to enhance the accuracy and reliability of generated responses. However, reliably attributing generated content to specific context segments, context attribution, remains challenging due to the computationally intensive nature of current methods, which often require extensive fine-tuning or human annotation. In this work, we introduce a novel Jensen-Shannon Divergence driven method to Attribute Response to Context (ARC-JSD), enabling efficient and accurate identification of essential context sentences without additional fine-tuning, gradient-calculation or surrogate modelling. Evaluations on a wide range of RAG benchmarks, such as TyDi QA, Hotpot QA, and Musique, using instruction-tuned LLMs in different scales demonstrate superior accuracy and significant computational efficiency improvements compared to the previous surrogate-based method. Furthermore, our mechanistic analysis reveals specific attention heads and multilayer perceptron (MLP) layers responsible for context attribution, providing valuable insights into the internal workings of RAG models and how they affect RAG behaviours. Our code is available at https://github.com/ruizheliUOA/ARC_JSD.
Authors: Jian Lan, Udo Schlegel, Tanveer Hannan, Gengyuan Zhang, Haokun Chen, Thomas Seidl
Abstract: Although Vision-Language Models (VLMs) have achieved remarkable success, the knowledge mechanisms underlying their social biases remain a black box, where fairness- and ethics-related problems harm certain groups of people in society. It is unknown to what extent VLMs yield gender and race bias in generative responses. In this paper, we conduct a systematic discovery of gender and race bias in state-of-the-art VLMs, focusing not only on surface-level responses but also on the internal probability distributions and hidden state dynamics. Our empirical analysis reveals three critical findings: 1) The Fairness Paradox: Models often generate fair text labels while maintaining highly skewed confidence scores (mis-calibration) toward specific social groups. 2) Layer-wise Fluctuation: Fairness knowledge is not uniformly distributed; it peaks in intermediate layers and undergoes substantial knowledge erosion in the final layers. 3) Residual Discrepancy: Within a single hidden layer, different residual streams carry conflicting social knowledge - some reinforcing fairness while others amplifying bias. Leveraging these insights, we propose RES-FAIR (RESidual Flow Adjustment for Inference Recalibration), a post-hoc framework that mitigates bias by localizing and projecting hidden states away from biased residual directions while amplifying fair components. Evaluations on PAIRS and SocialCounterfactuals datasets demonstrate that our discovery-based approach significantly improves response fairness and confidence calibration without compromising general reasoning abilities. Our work provides a new lens for understanding how multi-modal models store and process sensitive social information.
Authors: Yu He, Yingxi Li, Colin White, Ellen Vitercik
Abstract: Large language models (LLMs) are deployed on increasingly complex tasks that require multi-step decision-making. Understanding their algorithmic reasoning abilities is therefore crucial. However, we lack a diagnostic benchmark for evaluating this capability. We propose data structures as a principled lens: as fundamental building blocks of algorithms, they naturally probe structural reasoning-the ability to understand and manipulate relationships such as order, hierarchy, and connectivity that underpin algorithmic reasoning. We introduce DSR-Bench, spanning 20 data structures, 35 operations, and 4,140 problem instances. DSR-Bench features hierarchical task organization, fully automated generation and evaluation, and fine-grained diagnostics. Evaluating 13 state-of-the-art LLMs reveals critical limitations: the top-performing model achieves only 0.46/1 on challenging instances. Three auxiliary probes targeting more realistic usages expose further weaknesses: models perform poorly on spatial data and context-rich scenarios, and they struggle to reason over their own code.
Authors: Wei Zhong, Manasa Bharadwaj, Yixiao Wang, Yipeng Ji, Chul Lee
Abstract: Speculative decoding (SD) is a widely adopted approach for accelerating inference in large language models (LLMs), particularly when the draft and target models are well aligned. However, state-of-the-art SD methods typically rely on tightly coupled, self-attention-based Transformer decoders, often augmented with auxiliary pooling or fusion layers. This coupling makes them increasingly complex and harder to generalize across different models. We present Budget EAGLE (Beagle), the first, to our knowledge, cross-attention-based Transformer decoder SD model that achieves performance on par with leading self-attention SD models (EAGLE-v2) while eliminating the need for pooling or auxiliary components, simplifying the architecture, improving training efficiency, and maintaining stable memory usage during training-time simulation. To enable effective training of this novel architecture, we propose Two-Stage Block-Attention Training, a new method that achieves training stability and convergence efficiency in block-level attention scenarios. Extensive experiments across multiple LLMs and datasets show that Beagle achieves competitive inference speedups and higher training efficiency than EAGLE-v2, offering a strong alternative for architectures in speculative decoding.
Authors: Simon Sinong Zhan, Qingyuan Wu, Philip Wang, Frank Yang, Xiangyu Shi, Chao Huang, Qi Zhu
Abstract: Offline-to-online deployment of reinforcement-learning (RL) agents must bridge two gaps: (1) the sim-to-real gap, where real systems add latency and other imperfections not present in simulation, and (2) the interaction gap, where policies trained purely offline face out-of-distribution states during online execution because gathering new interaction data is costly or risky. Agents therefore have to generalize from static, delay-free datasets to dynamic, delay-prone environments. Standard offline RL learns from delay-free logs yet must act under delays that break the Markov assumption and hurt performance. We introduce DT-CORL (Delay-Transformer belief policy Constrained Offline RL), an offline-RL framework built to cope with delayed dynamics at deployment. DT-CORL (i) produces delay-robust actions with a transformer-based belief predictor even though it never sees delayed observations during training, and (ii) is markedly more sample-efficient than na\"ive history-augmentation baselines. Experiments on D4RL benchmarks with several delay settings show that DT-CORL consistently outperforms both history-augmentation and vanilla belief-based methods, narrowing the sim-to-real latency gap while preserving data efficiency.
Authors: Adrian de Wynter
Abstract: An evaluator, such as an LLM-as-a-judge, is trustworthy when there exists some agreed-upon way to measure its performance as a labeller. Traditional approaches either rely on testing the evaluator against references or assume that it `knows' somehow the correct labelling. Both approaches fail when references are unavailable: the former requires data, and the latter is an assumption, not evidence. To address this, we introduce the `No-Data Algorithm', which provably establishes trust in an evaluator without requiring any labelled data. Our algorithm works by successively posing challenges to said evaluator. We prove that after $r$ challenge rounds, it accepts an evaluator which knows the correct labels with probability $ \geq 1 - (1/4)^r$, and reliably flags untrustworthy ones. We present formal proofs of correctness, empirical tests, and applications to assessing trust in LLMs-as-judges for low-resource language labelling. Our work enables scientifically-grounded evaluator trust in low-data domains, addressing a critical bottleneck for scalable, trustworthy LLM deployment.
Authors: Shangpin Peng, Weinong Wang, Zhuotao Tian, Senqiao Yang, Xing Wu, Haotian Xu, Chengquan Zhang, Takashi Isobe, Baotian Hu, Min Zhang
Abstract: Direct Preference Optimization (DPO) has emerged as a cornerstone of reinforcement learning from human feedback (RLHF) due to its simplicity and efficiency. However, existing DPO-based methods typically treat all preference pairs equally, overlooking substantial variations in data quality and learning difficulty, which leads to inefficient data utilization and suboptimal performance. To address this limitation, we propose Uni-DPO, a unified dynamic preference optimization framework that jointly considers (a) the inherent quality of preference pairs and (b) the model's evolving performance during training. By adaptively reweighting samples based on both factors, Uni-DPO enables more effective use of preference data and achieves superior performance. Extensive experiments across models and benchmarks demonstrate the effectiveness and generalization of Uni-DPO. On textual tasks, Gemma-2-9B-IT fine-tuned with Uni-DPO surpasses the leading LLM, Claude 3 Opus, by 6.7 points on Arena-Hard. On mathematical and multimodal tasks, Uni-DPO consistently outperforms baseline methods across all benchmarks, providing strong empirical evidence of its effectiveness and robustness.
Authors: Chuan He, Zhaosong Lu, Defeng Sun, Zhanwang Deng
Abstract: In this paper, we propose practical normalized stochastic first-order methods with Polyak momentum, multi-extrapolated momentum, and recursive momentum for solving unconstrained optimization problems. These methods employ dynamically updated algorithmic parameters and do not require explicit knowledge of problem-dependent quantities such as the Lipschitz constant or noise bound. We establish first-order oracle complexity results for finding approximate stochastic stationary points under heavy-tailed noise and weakly average smoothness conditions -- both of which are weaker than the commonly used bounded variance and mean-squared smoothness assumptions. Our complexity bounds either improve upon or match the best-known results in the literature. Numerical experiments are presented to demonstrate the practical effectiveness of the proposed methods.
Authors: Seungoh Han, Jaehoon Jang, Hyunsu Kim, Jaeheung Surh, Junhyung Kwak, Hyowon Ha, Kyungdon Joo
Abstract: We introduce LighthouseGS, a practical novel view synthesis framework based on 3D Gaussian Splatting that utilizes simple panorama-style captures from a single mobile device. While convenient, this rotation-dominant motion and narrow baseline make accurate camera pose and 3D point estimation challenging, especially in textureless indoor scenes. To address these challenges, LighthouseGS leverages rough geometric priors, such as mobile device camera poses and monocular depth estimation, and utilizes indoor planar structures. Specifically, we propose a new initialization method called plane scaffold assembly to generate consistent 3D points on these structures, followed by a stable pruning strategy to enhance geometry and optimization stability. Additionally, we present geometric and photometric corrections to resolve inconsistencies from motion drift and auto-exposure in mobile devices. Tested on real and synthetic indoor scenes, LighthouseGS delivers photorealistic rendering, outperforming state-of-the-art methods and enabling applications like panoramic view synthesis and object placement. Project page: https://vision3d-lab.github.io/lighthousegs/
Authors: Jeffrey Joan Sam, Janhavi Sathe, Nikhil Chigali, Naman Gupta, Radhey Ruparel, Yicheng Jiang, Janmajay Singh, James W. Berck, Arko Barman
Abstract: Spacecraft deployed in outer space are routinely subjected to various forms of damage due to exposure to hazardous environments. In addition, there are significant risks to the subsequent process of in-space repairs through human extravehicular activity or robotic manipulation, incurring substantial operational costs. Recent developments in image segmentation could enable the development of reliable and cost-effective autonomous inspection systems. While these models often require large amounts of training data to achieve satisfactory results, publicly available annotated spacecraft segmentation data are very scarce. Here, we present a new dataset of nearly 64k annotated spacecraft images that was created using real spacecraft models, superimposed on a mixture of real and synthetic backgrounds generated using NASA's TTALOS pipeline. To mimic camera distortions and noise in real-world image acquisition, we also added different types of noise and distortion to the images. Our dataset includes images with several real-world challenges, including noise, camera distortions, glare, varying lighting conditions, varying field of view, partial spacecraft visibility, brightly-lit city backgrounds, densely patterned and confounding backgrounds, aurora borealis, and a wide variety of spacecraft geometries. Finally, we finetuned YOLOv8 and YOLOv11 models for spacecraft segmentation to generate performance benchmarks for the dataset under well-defined hardware and inference time constraints to mimic real-world image segmentation challenges for real-time onboard applications in space on NASA's inspector spacecraft. The resulting models, when tested under these constraints, achieved a Dice score of 0.92, Hausdorff distance of 0.69, and an inference time of about 0.5 second. The dataset and models for performance benchmark are available at https://github.com/RiceD2KLab/SWiM.
Authors: Linghao Zhu, Yiran Guan, Dingkang Liang, Jianzhong Ju, Zhenbo Luo, Bin Qin, Jian Luan, Yuliang Liu, Xiang Bai
Abstract: Reinforcement learning (RL) has emerged as an effective post-training paradigm for enhancing the reasoning capabilities of multimodal large language model (MLLM). However, current RL pipelines often suffer from training inefficiencies caused by two underexplored issues: Advantage Collapsing, where most advantages in a batch concentrate near zero, and Rollout Silencing, where the proportion of rollouts contributing non-zero gradients diminishes over time. These issues lead to suboptimal gradient updates and hinder long-term learning efficiency. To address these issues, we propose Shuffle-R1, a simple yet principled framework that improves RL fine-tuning efficiency by dynamically restructuring trajectory sampling and batch composition. It introduces (1) Pairwise Trajectory Sampling, which selects high-contrast trajectories with large advantages to improve gradient signal quality, and (2) Advantage-based Trajectory Shuffle, which increases exposure of valuable rollouts through informed batch reshuffling. Experiments across multiple reasoning benchmarks show that our framework consistently outperforms strong RL baselines with minimal overhead. These results highlight the importance of data-centric adaptations for more efficient RL training in MLLM.
Authors: Fan Zhang, Zebang Cheng, Chong Deng, Haoxuan Li, Zheng Lian, Qian Chen, Huadai Liu, Wen Wang, Yi-Fan Zhang, Renrui Zhang, Ziyu Guo, Zhihong Zhu, Hao Wu, Haixin Wang, Yefeng Zheng, Xiaojiang Peng, Xian Wu, Kun Wang, Xiangang Li, Jieping Ye, Pheng-Ann Heng
Abstract: Recent advances in multimodal large language models (MLLMs) have catalyzed transformative progress in affective computing, enabling models to exhibit emergent emotional intelligence. Despite substantial methodological progress, current emotional benchmarks remain limited, as it is still unknown: (a) the generalization abilities of MLLMs across distinct scenarios, and (b) their reasoning capabilities to identify the triggering factors behind emotional states. To bridge these gaps, we present \textbf{MME-Emotion}, a systematic benchmark that assesses both emotional understanding and reasoning capabilities of MLLMs, enjoying \textit{scalable capacity}, \textit{diverse settings}, and \textit{unified protocols}. As the largest emotional intelligence benchmark for MLLMs, MME-Emotion contains over 6,000 curated video clips with task-specific questioning-answering (QA) pairs, spanning broad scenarios to formulate eight emotional tasks. It further incorporates a holistic evaluation suite with hybrid metrics for emotion recognition and reasoning, analyzed through a multi-agent system framework. Through a rigorous evaluation of 20 advanced MLLMs, we uncover both their strengths and limitations, yielding several key insights: \ding{182} Current MLLMs exhibit unsatisfactory emotional intelligence, with the best-performing model achieving only $39.3\%$ recognition score and $56.0\%$ Chain-of-Thought (CoT) score on our benchmark. \ding{183} Generalist models (\emph{e.g.}, Gemini-2.5-Pro) derive emotional intelligence from generalized multimodal understanding capabilities, while specialist models (\emph{e.g.}, R1-Omni) can achieve comparable performance through domain-specific post-training adaptation. By introducing MME-Emotion, we hope that it can serve as a foundation for advancing MLLMs' emotional intelligence in the future.
Authors: Amine Lbath, Massih-Reza Amini, Aurelien Delaitre, Vadim Okun
Abstract: The increasing complexity of software systems and the sophistication of cyber-attacks have underscored the need for reliable automated software vulnerability detection. Data-driven approaches using deep learning models show promise but critically depend on the availability of large, accurately labeled datasets. Yet existing datasets either suffer from noisy labels, limited vulnerability coverage, or fail to reflect vulnerabilities as they occur in real-world software. This also limits large-scale benchmarking of such solutions. Automated vulnerability injection provides a way to address these limitations, but existing techniques remain limited in coverage, contextual fidelity, or injection success. In this paper, we present AVIATOR, the first AI-agentic vulnerability injection framework. AVIATOR decomposes vulnerability injection into a coordinated workflow of specialized AI agents, tool-based analysis, and iterative self-correction, explicitly mirroring expert reasoning. It integrates RAG and lightweight LoRA-based fine-tuning to produce realistic, category-specific vulnerabilities without relying on handcrafted patterns. Across three benchmarks, AVIATOR achieves high injection fidelity (91-95%) surpassing existing injection techniques in both accuracy and vulnerability coverage. When used for data augmentation to train deep learning-based vulnerability detection (DLVD) models, AVIATOR provides the strongest downstream gains in vulnerability detection. Across models and base datasets, AVIATOR improves average F1 scores by +22% over no augmentation, +25% over VGX, holding the prior best injection success rate, and +3% over VulScribeR, the prior state-of-the-art LLM-based injection model, with +7% higher recall and no precision loss. Its augmented data exhibits the lowest distributional distortion and scales efficiently with <2% syntax rejection at 4.3x lower cost than VulScribeR.
Authors: Qizhou Wang, Hanxun Huang, Guansong Pang, Sarah Erfani, Christopher Leckie
Abstract: Speech synthesis systems can now produce highly realistic vocalisations that pose significant authenticity challenges. Despite substantial progress in deepfake detection models, their real-world effectiveness is often undermined by evolving distribution shifts between training and test data, driven by the complexity of human speech and the rapid evolution of synthesis systems. Existing datasets suffer from limited real speech diversity, insufficient coverage of recent synthesis systems, and heterogeneous mixtures of deepfake sources, which hinder systematic evaluation and open-world model training. To address these issues, we introduce AUDETER (AUdio DEepfake TEst Range), a large-scale and highly diverse deepfake audio dataset comprising over 4,500 hours of synthetic audio generated by 11 recent TTS models and 10 vocoders, totalling 3 million clips. We further observe that most existing detectors default to binary supervised training, which can induce negative transfer across synthesis sources when the training data contains highly diverse deepfake patterns, impacting overall generalisation. As a complementary contribution, we propose an effective curriculum-learning-based approach to mitigate this effect. Extensive experiments show that existing detection models struggle to generalise to novel deepfakes and human speech in AUDETER, whereas XLR-based detectors trained on AUDETER achieve strong cross-domain performance across multiple benchmarks, achieving an EER of 1.87% on In-the-Wild. AUDETER is available on GitHub.
Authors: Long Li, Zhijian Zhou, Jiaran Hao, Jason Klein Liu, Yanting Miao, Wei Pang, Xiaoyu Tan, Wei Chu, Zhe Wang, Shirui Pan, Chao Qu, Yuan Qi
Abstract: A central paradox in fine-tuning Large Language Models (LLMs) with Reinforcement Learning with Verifiable Reward (RLVR) is the frequent degradation of multi-attempt performance (Pass@k) despite improvements in single-attempt accuracy (Pass@1). This is often accompanied by catastrophic forgetting, where models lose previously acquired skills. While various methods have been proposed, the choice and function of the divergence term have been surprisingly unexamined as a proactive solution. We argue that standard RLVR objectives -- both those using the mode-seeking reverse KL-divergence and those forgoing a divergence term entirely -- lack a crucial mechanism for knowledge retention. The reverse-KL actively accelerates this decay by narrowing the policy, while its absence provides no safeguard against the model drifting from its diverse knowledge base. We propose a fundamental shift in perspective: using the divergence term itself as the solution. Our framework, Diversity-Preserving Hybrid RL (DPH-RL), leverages mass-covering f-divergences (like forward-KL and JS-divergence) to function as a rehearsal mechanism. By continuously referencing the initial policy, this approach forces the model to maintain broad solution coverage. Extensive experiments on math and SQL generation demonstrate that DPH-RL not only resolves the Pass@k degradation but improves both Pass@1 and Pass@k in- and out-of-domain. Additionally, DPH-RL is more training-efficient because it computes f-divergence using generator functions, requiring only sampling from the initial policy and no online reference model. Our work highlights a crucial, overlooked axis for improving RLVR, demonstrating that the proper selection of a divergence measure is a powerful tool for building more general and diverse reasoning models.
Authors: Bingxin Xu, Zhen Dong, Oussama Elachqar, Yuzhang Shang
Abstract: Large language models require massive memory footprints, severely limiting deployment on consumer hardware. Quantization reduces memory through lower numerical precision, but extreme 2-bit quantization suffers from catastrophic performance loss due to outliers in activations. Rotation-based methods such as QuIP and QuaRot apply orthogonal transforms to eliminate outliers before quantization, using computational invariance: $\mathbf{y} = \mathbf{Wx} = (\mathbf{WQ}^T)(\mathbf{Qx})$ for orthogonal $\mathbf{Q}$. However, these methods use fixed transforms--Hadamard matrices achieving optimal worst-case coherence $\mu = 1/\sqrt{n}$--that cannot adapt to specific weight distributions. We identify that different transformer layers exhibit distinct outlier patterns, motivating layer-adaptive rotations rather than one-size-fits-all approaches. In this work, we propose ButterflyQuant, which replaces Hadamard rotations with learnable butterfly transforms parameterized by continuous Givens rotation angles. Unlike Hadamard's discrete $\{+1, -1\}$ entries that are non-differentiable and thus prohibit gradient-based learning, butterfly transforms' continuous parameterization enables smooth optimization while guaranteeing orthogonality by construction. This orthogonal constraint ensures theoretical guarantees in outlier suppression while achieving $O(n \log n)$ computational complexity with only $\frac{n \log n}{2}$ learnable parameters. We further introduce a uniformity regularization on post-transformation activations to promote smoother distributions amenable to quantization. Learning requires only 128 calibration samples and converges in minutes on a single GPU.
Authors: Hanbit Oh, Masaki Murooka, Tomohiro Motoda, Ryoichi Nakajo, Yukiyasu Domae
Abstract: Imitation learning is a promising paradigm for training robot agents; however, standard approaches typically require substantial data acquisition -- via numerous demonstrations or random exploration -- to ensure reliable performance. Although exploration reduces human effort, it lacks safety guarantees and often results in frequent collisions -- particularly in clearance-limited tasks (e.g., peg-in-hole) -- thereby, necessitating manual environmental resets and imposing additional human burden. This study proposes Self-Augmented Robot Trajectory (SART), a framework that enables policy learning from a single human demonstration, while safely expanding the dataset through autonomous augmentation. SART consists of two stages: (1) human teaching only once, where a single demonstration is provided and precision boundaries -- represented as spheres around key waypoints -- are annotated, followed by one environment reset; (2) robot self-augmentation, where the robot generates diverse, collision-free trajectories within these boundaries and reconnects to the original demonstration. This design improves the data collection efficiency by minimizing human effort while ensuring safety. Extensive evaluations in simulation and real-world manipulation tasks show that SART achieves substantially higher success rates than policies trained solely on human-collected demonstrations. Video results available at https://sites.google.com/view/sart-il .
Authors: Adrian de Wynter
Abstract: In-context learning (ICL) allows some autoregressive models to solve tasks via next-token prediction and without needing further training. This has led to claims about these model's ability to solve (learn) unseen tasks with only a few shots (exemplars) in the prompt. However, deduction does not always imply learning, as ICL does not explicitly encode a given observation. Instead, the models rely on their prior knowledge and the exemplars given, if any. We argue that, mathematically, ICL fits the definition of learning; however, its full characterisation requires empirical work. We then carry out a large-scale analysis of ICL ablating out or accounting for memorisation, pretraining, distributional shifts, and prompting style and phrasing. We find that, empirically, ICL is limited in its ability to learn and generalise to unseen tasks. Namely, in the limit where exemplars become more numerous, accuracy is insensitive to exemplar distribution, model, prompt style, and the input's linguistic features. Instead, it deduces patterns from regularities in the prompt, which leads to distributional sensitivity, especially in prompting styles such as chain-of-thought. Given the varied accuracies and on formally similar tasks, we conclude that autoregression's ad-hoc encoding is not a robust mechanism for learning, and suggests limited all-purpose generalisability.
Authors: Xiaobo Xing, Wei Yuan, Tong Chen, Quoc Viet Hung Nguyen, Xiangliang Zhang, Hongzhi Yin
Abstract: Modeling semantic and structural information from tabular data remains a core challenge for effective table understanding. Existing Table-as-Text approaches flatten tables for large language models (LLMs), but lose crucial structural cues, while Table-as-Image methods preserve structure yet struggle with precise semantics. Recent Table-as-Multimodality strategies attempt to combine textual and visual views, but they (1) statically process both modalities for every query-table pair within large multimodal LLMs (MLLMs), inevitably introducing redundancy and even conflicts, and (2) depend on costly fine-tuning of MLLMs. In light of this, we propose TableDART, a training-efficient framework that integrates multimodal views by reusing pretrained single-modality models. TableDART introduces a lightweight 2.59M-parameter MLP gating network that dynamically selects the optimal path (Text-only, Image-only, or Fusion) for each table-query pair, reducing redundancy and avoiding conflicts that arise when textual and visual views of the same table provide inconsistent cues. By routing to the most appropriate view, our framework improves both accuracy and efficiency. In addition, we propose a novel agent to mediate cross-modal knowledge integration by analyzing outputs from text- and image-based models, either selecting the best result or synthesizing a new answer through reasoning. This design avoids the prohibitive costs of full MLLM fine-tuning. Extensive experiments on seven benchmarks show that TableDART establishes new state-of-the-art performance among open-source models, surpassing the strongest baseline by an average of 4.02%. The code is available at: https://github.com/xiaobo-xing/TableDART.
Authors: Junhyeok Lee, Helin Wang, Yaohan Guan, Thomas Thebaud, Laureano Moro-Velazquez, Jes\'us Villalba, Najim Dehak
Abstract: We introduce MaskVCT, a zero-shot voice conversion (VC) model that offers multi-factor controllability through multiple classifier-free guidances (CFGs). While previous VC models rely on a fixed conditioning scheme, MaskVCT integrates diverse conditions in a single model. To further enhance robustness and control, the model can leverage continuous or quantized linguistic features to enhance intelligibility and speaker similarity, and can use or omit pitch contour to control prosody. These choices allow users to seamlessly balance speaker identity, linguistic content, and prosodic factors in a zero-shot VC setting. Extensive experiments demonstrate that MaskVCT achieves the best target speaker and accent similarities while obtaining competitive word and character error rates compared to existing baselines. Audio samples are available at https://maskvct.github.io/.
Authors: Yinuo Wang, Yuanyang Qi, Jinzhao Zhou, Pengxiang Meng, Xiaowen Tao
Abstract: End-to-end reinforcement learning (RL) for humanoid locomotion is appealing for its compact perception-action mapping, yet practical policies often suffer from training instability, inefficient feature fusion, and high actuation cost. We present HuMam, a state-centric end-to-end RL framework that employs a single-layer Mamba encoder to fuse robot-centric states with oriented footstep targets and a continuous phase clock. The policy outputs joint position targets tracked by a low-level PD loop and is optimized with PPO. A concise six-term reward balances contact quality, swing smoothness, foot placement, posture, and body stability while implicitly promoting energy saving. On the JVRC-1 humanoid in mc-mujoco, HuMam consistently improves learning efficiency, training stability, and overall task performance over a strong feedforward baseline, while reducing power consumption and torque peaks. To our knowledge, this is the first end-to-end humanoid RL controller that adopts Mamba as the fusion backbone, demonstrating tangible gains in efficiency, stability, and control economy.
Authors: Yifan Wang, Mayank Jobanputra, Ji-Ung Lee, Soyoung Oh, Isabel Valera, Vera Demberg
Abstract: Natural language processing (NLP) models often replicate or amplify social bias from training data, raising concerns about fairness. At the same time, their black-box nature makes it difficult for users to recognize biased predictions and for developers to effectively mitigate them. While some studies suggest that input-based explanations can help detect and mitigate bias, others question their reliability in ensuring fairness. Existing research on explainability in fair NLP has been predominantly qualitative, with limited large-scale quantitative analysis. In this work, we conduct the first systematic study of the relationship between explainability and fairness in hate speech detection, focusing on both encoder- and decoder-only models. We examine three key dimensions: (1) identifying biased predictions, (2) selecting fair models, and (3) mitigating bias during model training. Our findings show that input-based explanations can effectively detect biased predictions and serve as useful supervision for reducing bias during training, but they are unreliable for selecting fair models among candidates.Our code is available at https://github.com/Ewanwong/fairness_x_explainability.
URLs: https://github.com/Ewanwong/fairness_x_explainability.
Authors: Zijian Wang, Xiaofei Zhang, Xin Zhang, Yukun Liu, Qiong Zhang
Abstract: Federated learning (FL) is increasingly adopted in domains like healthcare, where data privacy is paramount. A fundamental challenge in these systems is statistical heterogeneity-the fact that data distributions vary significantly across clients (e.g., different hospitals may treat distinct patient demographics). While current FL algorithms focus on aggregating model updates from these heterogeneous clients, the potential of the central server remains under-explored. This paper is motivated by a healthcare scenario: could a central server not only coordinate model training but also guide a new patient to the hospital best equipped for their specific condition? We generalize this idea to propose a novel paradigm for FL systems where the server actively guides the allocation of new tasks or queries to the most appropriate client. To enable this, we introduce a density ratio model and empirical likelihood-based framework that simultaneously addresses two goals: (1) learning effective local models on each client, and (2) finding the best matching client for a new query. Empirical results demonstrate the framework's effectiveness on benchmark datasets, showing improvements in both model accuracy and the precision of client guidance compared to standard FL approaches. This work opens a new direction for building more intelligent and resource-efficient FL systems that leverage heterogeneity as a feature, not just a bug. Code is available at https://github.com/zijianwang0510/FedDRM.git.
Authors: Lin Long, Changdae Oh, Seongheon Park, Sharon Li
Abstract: Large vision-language models (LVLMs) achieve strong performance on multimodal tasks, yet they often default to their language prior (LP) -- memorized textual patterns from pre-training while under-utilizing visual evidence. Prior analyses of LP mostly rely on input-output probing, which fails to reveal the internal mechanisms governing when and how vision influences model behavior. To address this gap, we present the first systematic analysis of language prior through the lens of chain-of-embedding, which examines the layer-wise representation dynamics within LVLMs. Our analysis reveals a universal phenomenon: each model exhibits a Visual Integration Point (VIP), a critical layer at which visual information begins to meaningfully reshape hidden representations and influence decoding for multimodal reasoning. Building on this observation, we introduce the Total Visual Integration (TVI) estimator, which aggregates representational discrepancy beyond the VIP to quantify how strongly visual query influences response generation. Across 60 model-dataset combinations spanning 10 contemporary LVLMs and 6 benchmarks, we demonstrate that VIP consistently emerges, and that TVI reliably predicts the strength of language prior. This offers a principled toolkit for diagnosing and understanding language prior in LVLMs.
Authors: Langqi Yang, Tianhang Zheng, Yixuan Chen, Kedong Xiu, Hao Zhou, Di Wang, Puning Zhao, Zhan Qin, Kui Ren
Abstract: The potential for large language models (LLMs) to generate harmful content poses a significant safety risk in their deployment. To address and assess this risk, the community has developed numerous harmfulness evaluation metrics and judges. However, the lack of a systematic benchmark for evaluating these metrics and judges undermines the credibility and consistency of LLM safety assessments. To bridge this gap, we introduce HarmMetric Eval, a comprehensive benchmark designed to support both overall and fine-grained evaluation of harmfulness metrics and judges. In HarmMetric Eval, we build a high-quality dataset of representative harmful prompts paired with highly diverse harmful model responses and non-harmful counterparts across multiple categories. We also propose a flexible scoring mechanism that rewards the metrics for correctly ranking harmful responses above non-harmful ones, which is applicable to almost all existing metrics and judges with varying output formats and scoring scales. Using HarmMetric Eval, we uncover a surprising finding by extensive experiments: Conventional reference-based metrics such as ROUGE and METEOR can outperform existing LLM-based judges in fine-grained harmfulness evaluation, challenging prevailing assumptions about LLMs'superiority in this domain. To reveal the reasons behind this finding, we provide a fine-grained analysis to explain the limitations of LLM-based judges on rating irrelevant or useless responses. Furthermore, we build a new harmfulness judge by incorporating the fine-grained criteria into its prompt template and leverage reference-based metrics to fine-tune its base LLM. The resulting judge demonstrates superior performance than all existing metrics and judges in evaluating harmful responses.
Authors: Michael Drolet, Firas Al-Hafez, Aditya Bhatt, Jan Peters, Oleg Arenz
Abstract: Discrete latent bottlenecks in variational autoencoders (VAEs) offer high bit efficiency and can be modeled with autoregressive discrete distributions, enabling parameter-efficient multimodal search with transformers. However, discrete random variables do not allow for exact differentiable parameterization; therefore, discrete VAEs typically rely on approximations, such as Gumbel-Softmax reparameterization or straight-through gradient estimates, or employ high-variance gradient-free methods such as REINFORCE that have had limited success on high-dimensional tasks such as image reconstruction. Inspired by popular techniques in policy search, we propose a training framework for discrete VAEs that leverages the natural gradient of a non-parametric encoder to update the parametric encoder without requiring reparameterization. Our method, combined with automatic step size adaptation and a transformer-based encoder, scales to challenging datasets such as ImageNet and outperforms both approximate reparameterization methods and quantization-based discrete autoencoders in reconstructing high-dimensional data from compact latent spaces.
Authors: Peter Holderrieth, Uriel Singer, Tommi Jaakkola, Ricky T. Q. Chen, Yaron Lipman, Brian Karrer
Abstract: The performance of flow matching and diffusion models can be greatly improved at inference time using reward alignment algorithms, yet efficiency remains a major limitation. While several algorithms were proposed, we demonstrate that a common bottleneck is the sampling method these algorithms rely on: many algorithms require to sample Markov transitions via SDE sampling, which is significantly less efficient and often less performant than ODE sampling. To remove this bottleneck, we introduce GLASS Flows, a new sampling paradigm that simulates a "flow matching model within a flow matching model" to sample Markov transitions. As we show in this work, this "inner" flow matching model can be retrieved from a pre-trained model without any re-training, combining the efficiency of ODEs with the stochastic evolution of SDEs. On large-scale text-to-image models, we show that GLASS Flows eliminate the trade-off between stochastic evolution and efficiency. Combined with Feynman-Kac Steering, GLASS Flows improve state-of-the-art performance in text-to-image generation, making it a simple, drop-in solution for inference-time scaling of flow and diffusion models.
Authors: Chi Zhang, Zehua Chen, Kaiwen Zheng, Jun Zhu
Abstract: Bridge models have recently been explored for speech enhancement tasks such as denoising, dereverberation, and super-resolution, while these efforts are typically confined to a single task or small-scale datasets, with constrained general speech restoration (GSR) capability at scale. In this work, we introduce VoiceBridge, a GSR system rooted in latent bridge models (LBMs), capable of reconstructing high-fidelity speech at full-band (\textit{i.e.,} 48~kHz) from various distortions. By compressing speech waveform into continuous latent representations, VoiceBridge models the~\textit{diverse LQ-to-HQ tasks} (namely, low-quality to high-quality) in GSR with~\textit{a single latent-to-latent generative process} backed by a scalable transformer architecture. To better inherit the advantages of bridge models from the data domain to the latent space, we present an energy-preserving variational autoencoder, enhancing the alignment between the waveform and latent space over varying energy levels. Furthermore, to address the difficulty of HQ reconstruction from distinctively different LQ priors, we propose a joint neural prior, uniformly alleviating the reconstruction burden of LBM. At last, considering the key requirement of GSR systems, human perceptual quality, a perceptually aware fine-tuning stage is designed to mitigate the cascading mismatch in generation while improving perceptual alignment. Extensive validation across in-domain and out-of-domain tasks and datasets (\textit{e.g.}, refining recent zero-shot speech and podcast generation results) demonstrates the superior performance of VoiceBridge. Demo samples can be visited at: https://VoiceBridge-demo.github.io/.
Authors: Vincent Grari, Tim Arni, Thibault Laugel, Sylvain Lamprier, James Zou, Marcin Detyniecki
Abstract: When used in high-stakes settings, AI systems are expected to produce decisions that are transparent, interpretable and auditable, a requirement increasingly expected by regulations. Decision trees such as CART provide clear and verifiable rules, but they are restricted to structured tabular data and cannot operate directly on unstructured inputs such as text. In practice, large language models (LLMs) are widely used for such data, yet prompting strategies such as chain-of-thought or prompt optimization still rely on free-form reasoning, limiting their ability to ensure trustworthy behaviors. We present the Agentic Classification Tree (ACT), which extends decision-tree methodology to unstructured inputs by formulating each split as a natural-language question, refined through impurity-based evaluation and LLM feedback via TextGrad. Experiments on text benchmarks show that ACT matches or surpasses prompting-based baselines while producing transparent and interpretable decision paths.
Authors: Yanming Li (PETSCRAFT), C\'edric Eichler (PETSCRAFT), Nicolas Anciaux (PETSCRAFT), Alexandra Bensamoun (UC3M), Lorena Gonzalez Manzano (UC3M), Seifeddine Ghozzi (ENSTA)
Abstract: We propose a system for marking sensitive or copyrighted texts to detect their use in fine-tuning large language models under black-box access with statistical guarantees. Our method builds digital ``marks'' using invisible Unicode characters organized into (``cue'', ``reply'') pairs. During an audit, prompts containing only ``cue'' fragments are issued to trigger regurgitation of the corresponding ``reply'', indicating document usage. To control false positives, we compare against held-out counterfactual marks and apply a ranking test, yielding a verifiable bound on the false positive rate. The approach is minimally invasive, scalable across many sources, robust to standard processing pipelines, and achieves high detection power even when marked data is a small fraction of the fine-tuning corpus.
Authors: Dechen Zhang, Junwei Su, Difan Zou
Abstract: The use of low-bit quantization has emerged as an indispensable technique for enabling the efficient training of large-scale models. Despite its widespread empirical success, a rigorous theoretical understanding of its impact on learning performance remains notably absent, even in the simplest linear regression setting. We present the first systematic theoretical study of this fundamental question, analyzing finite-step stochastic gradient descent (SGD) for high-dimensional linear regression under a comprehensive range of quantization targets: data, label, parameter, activation, and gradient. Our novel analytical framework establishes precise algorithm-dependent and data-dependent excess risk bounds that characterize how different quantization affects learning: parameter, activation, and gradient quantization amplify noise during training; data quantization distorts the data spectrum; and data quantization introduces additional approximation error. Crucially, we distinguish the effects of two quantization schemes: we prove that for additive quantization (with constant quantization steps), the noise amplification benefits from a suppression effect scaled by the batch size, while multiplicative quantization (with input-dependent quantization steps) largely preserves the spectral structure, thereby reducing the spectral distortion. Furthermore, under common polynomial-decay data spectra, we quantitatively compare the risks of multiplicative and additive quantization, drawing a parallel to the comparison between FP and integer quantization methods. Our theory provides a powerful lens to characterize how quantization shapes the learning dynamics of optimization algorithms, paving the way to further explore learning theory under practical hardware constraints.
Authors: Beiya Dai, Yuliang Liu, Daozheng Xue, Yunchong Song, Qipeng Guo, Kai Chen, Xinbing Wang, Bowen Zhou, Zhouhan Lin
Abstract: We propose ContextLM, a framework that implicitly learns multi-token prediction by augmenting standard pretraining with an intrinsic next-context prediction objective. ContextLM builds a language model on top of context embeddings that span multiple tokens, enabling better next-token prediction by predicting the next context. Our model is fully compatible with standard autoregressive, token-by-token evaluation paradigms (e.g., perplexity). Extensive experiments with GPT-2 and Pythia backbones (up to 1.5B parameters and 300B training tokens) reveal that ContextLM shifts the Pareto frontier of scaling laws, exhibiting superior efficiency in parameters, training tokens, and FLOPs. Our results show that ContextLM could already achieve the baseline perplexity using 39\% fewer parameters and demonstrates robust generalization improvements on extensive downstream tasks under equivalent parameter counts.
Authors: Ruiyi Yang, Hao Xue, Imran Razzak, Hakim Hacid, Flora D. Salim
Abstract: Retrieval-augmented generation (RAG) remains brittle on multi-step questions and heterogeneous evidence sources, trading accuracy against latency and token/tool budgets. This paper introduces RELOOP, a structure aware framework using Hierarchical Sequence (HSEQ) that (i) linearize documents, tables, and knowledge graphs into a reversible hierarchical sequence with lightweight structural tags, and (ii) perform structure-aware iteration to collect just-enough evidence before answer synthesis. A Head Agent provides guidance that leads retrieval, while an Iteration Agent selects and expands HSeq via structure-respecting actions (e.g., parent/child hops, table row/column neighbors, KG relations); Finally the head agent composes canonicalized evidence to genearte the final answer, with an optional refinement loop to resolve detected contradictions. Experiments on HotpotQA (text), HybridQA/TAT-QA (table+text), and MetaQA (KG) show consistent EM/F1 gains over strong single-pass, multi-hop, and agentic RAG baselines with high efficiency. Besides, RELOOP exhibits three key advantages: (1) a format-agnostic unification that enables a single policy to operate across text, tables, and KGs without per-dataset specialization; (2) \textbf{guided, budget-aware iteration} that reduces unnecessary hops, tool calls, and tokens while preserving accuracy; and (3) evidence canonicalization for reliable QA, improving answers consistency and auditability.
Authors: Ljupcho Milosheski, Kuon Akiyama, Bla\v{z} Bertalani\v{c}, Jernej Hribar, Ryoichi Shinkuma
Abstract: The growing number of smart devices supporting bandwidth-intensive and latency-sensitive applications, such as real-time video analytics, smart sensing, Extended Reality (XR), etc., necessitates reliable wireless connectivity in indoor environments. In such environments, accurate design of Radio Environment Maps (REMs) enables adaptive wireless network planning and optimization of Access Point (AP) placement. However, generating realistic REMs remains difficult due to the variability of indoor environments and the limitations of existing modeling approaches, which often rely on simplified layouts or fully synthetic data. These challenges are further amplified by the adoption of next-generation Wi-Fi standards, which operate at higher frequencies and suffer from limited range and wall penetration. To support the efforts in addressing these challenges, we collected a dataset that combines high-resolution 3D LiDAR scans with Wi-Fi RSSI measurements collected across 20 setups in a multi-room indoor environment. The dataset includes two measurement scenarios, the first without human presence in the environment, and the second with human presence, enabling the development and validation of REM estimation models that incorporate physical geometry and environmental dynamics. The described dataset supports research in data-driven wireless modeling and the development of high-capacity indoor communication networks.
Authors: Michelle Huang, Violeta J. Rodriguez, Koustuv Saha, Tal August
Abstract: Limited English proficiency (LEP) patients in the U.S. face systemic barriers to healthcare beyond language and interpreter access, encompassing procedural and institutional constraints. AI advances may support communication and care through on-demand translation and visit preparation, but also risk exacerbating existing inequalities. We conducted storyboard-driven interviews with 14 patient navigators to explore how AI could shape care experiences for Spanish-speaking LEP individuals. We identified tensions around linguistic and cultural misunderstandings, privacy concerns, and opportunities and risks for AI to augment care workflows. Participants highlighted structural factors that can undermine trust in AI systems, including sensitive information disclosure, unstable technology access, and low literacy. While AI tools can potentially alleviate social barriers and institutional constraints, there are risks of misinformation and reducing human-to-human interactions. Our findings contribute AI design considerations that support LEP patients and care teams via rapport-building, educational and language support, and minimizing disruptions to existing practices.
Authors: Yudong Yang, Xuezhen Zhang, Zhifeng Han, Siyin Wang, Jimin Zhuang, Zengrui Jin, Jing Shao, Guangzhi Sun, Chao Zhang
Abstract: Recent progress in LLMs has enabled understanding of audio signals, but has also exposed new safety risks arising from complex audio inputs that are inadequately handled by current safeguards. We introduce SACRED-Bench (Speech-Audio Composition for RED-teaming) to evaluate the robustness of LLMs under complex audio-based attacks. Unlike existing perturbation-based methods that rely on noise optimization or white-box access, SACRED-Bench exploits speech-audio composition to enable effective black-box attacks. SACRED-Bench adopts three composition mechanisms: (a) overlap of harmful and benign speech, (b) mixture of benign speech with harmful non-speech audio, and (c) multi-speaker dialogue. These mechanisms focus on evaluating safety in settings where benign and harmful intents co-occur within a single auditory scene. Moreover, questions in SACRED-Bench are designed to implicitly refer to content in the audio, such that no explicit harmful information appears in the text prompt alone. Experiments demonstrate that even Gemini 2.5 Pro, a state-of-the-art proprietary LLM with safety guardrails fully enabled, still exhibits a 66% attack success rate. To bridge this gap, we propose SALMONN-Guard, the first guard model that jointly inspects speech, audio, and text for safety judgments, reducing the attack success rate to 20%. Our results highlight the need for audio-aware defenses to ensure the safety of multimodal LLMs. The dataset and SALMONN-Guard checkpoints can be found at https://huggingface.co/datasets/tsinghua-ee/SACRED-Bench.
URLs: https://huggingface.co/datasets/tsinghua-ee/SACRED-Bench.
Authors: Ryan Singh, Austin Hamilton, Amanda White, Michael Wise, Ibrahim Yousif, Arthur Carvalho, Zhe Shan, Reza Abrisham Baf, Mohammad Mayyas, Lora A. Cavuoto, Fadel M. Megahed
Abstract: Ensuring worker safety remains a critical challenge in modern manufacturing environments. Industry 5.0 reorients the prevailing manufacturing paradigm toward more human-centric operations. Using a design science research methodology, we identify three essential requirements for next-generation safety training systems: high accuracy, low latency, and low cost. We introduce a multimodal chatbot powered by large language models that meets these design requirements. The chatbot uses retrieval-augmented generation to ground its responses in curated regulatory and technical documentation. To evaluate our solution, we developed a domain-specific benchmark of expert-validated question and answer pairs for three representative machines: a Bridgeport manual mill, a Haas TL-1 CNC lathe, and a Universal Robots UR5e collaborative robot. We tested 24 RAG configurations using a full-factorial design and assessed them with automated evaluations of correctness, latency, and cost. Our top 2 configurations were then evaluated by ten industry experts and academic researchers. Our results show that retrieval strategy and model configuration have a significant impact on performance. The top configuration, selected for chatbot deployment, achieved an accuracy of 86.66%, an average cost of $0.005 per query, and an average end-to-end latency of 10.04 seconds. This latency is practical for delivering a complete safety instruction and is measured from query submission to full instruction delivery rather than generation onset. Overall, our work provides three contributions: an open-source, domain-grounded safety training chatbot; a validated benchmark for evaluating AI-assisted safety instruction; and a systematic methodology for designing and assessing AI-enabled instructional and immersive safety training systems for Industry 5.0 environments.
Authors: Chengdong Ma, Th\'eo Tao Zhaowei, Pengyu Li, Minghao Liu, Haojun Chen, Zihao Mao, Yuan Cheng, Yuan Qi, Yaodong Yang
Abstract: Since Isaac Newton first studied the Kissing Number Problem in 1694, determining the maximal number of non-overlapping spheres around a central sphere has remained a fundamental challenge. This problem is the local analogue of Hilbert's 18th problem, bridging geometry, number theory, and information theory. Although significant progress has been made through lattices and codes, the irregularities of high-dimensional geometry, dimensional structure variability, and combinatorial explosion beyond Go limit the scalability and generality of existing methods. Here we model the problem as a two-player matrix completion game and train the reinforcement learning system, PackingStar, to play the games. The matrix entries represent pairwise cosines of sphere center vectors. One player fills entries while another corrects suboptimal ones to improve exploration quality, cooperatively maximizing the matrix size, corresponding to the kissing number. These matrices are decomposed into representative substructures, providing diverse bases and structural constraints that steer subsequent games and make extremely large spaces tractable. PackingStar surpasses records from dimensions 25 to 31 and sets new lower bounds for generalized kissing numbers under various angular constraints. It achieves the first breakthrough beyond rational structures from 1971 in 13 dimensions and discovers over 6000 new structures in other dimensions. Notably, some configurations challenge long-held antipodal paradigms, revealing algebraic correspondences with finite simple groups as well as geometric relationships across dimensions. Inspired by these patterns, humans devised further improved constructions. These results demonstrate AI's power to explore high-dimensional spaces beyond human intuition via extreme-scale reinforcement learning and open new pathways for the Kissing Number Problem and broader geometry research.
Authors: Dezhi Ran, Shuxiao Xie, Mingfang Ji, Anmin Liu, Mengzhou Wu, Yuan Cao, Yuzhe Guo, Hao Yu, Linyi Li, Yitao Hu, Wei Yang, Tao Xie
Abstract: High-performance GPU kernels are critical for efficient LLM serving, yet their optimization remains a bottleneck requiring deep system expertise. While code LLMs show promise in generating functionally correct code, kernel optimization is intrinsically a search problem over a vast optimization space. The fundamental mismatch prevents existing LLM agents from efficiently exploring the optimization space for diverse hardware and compute patterns. To bridge the gap, we present KernelBand, a framework that formulates kernel optimization as a Multi-Armed Bandit (MAB) problem, explicitly balancing exploration and exploitation to unlock the potential of code LLMs. To navigate the infinite arm space of optimization strategies applied to candidate kernels, we design two key mechanisms: a hardware-aware pruning strategy via profiling bounds and a trace-driven clustering algorithm that leverages Lipschitz continuity. Theoretically, we prove that KernelBand reduces the regret bound to depend on the compact covering number of runtime clusters, ensuring sample-efficient discovery of high-performance kernels. Extensive experiments on TritonBench-G with three GPU architectures and four code LLMs show that KernelBand consistently and substantially outperforms state-of-the-art methods with over 33% average improvement.
Authors: Seungjun Yu, Seonho Lee, Namho Kim, Jaeyo Shin, Junsung Park, Wonjeong Ryu, Raehyuk Jung, Hyunjung Shim
Abstract: Recent advancements in multimodal large language models (MLLMs) have shown strong understanding of driving scenes, drawing interest in their application to autonomous driving. However, high-level reasoning in safety-critical scenarios, where avoiding one traffic risk can create another, remains a major challenge. Such reasoning is often infeasible with only a single front view and requires a comprehensive view of the environment, which we achieve through multi-view inputs. We define Safety-Critical Reasoning as a new task that leverages multi-view inputs to address this challenge. Then, we distill Safety-Critical Reasoning into two stages: first resolve the immediate risk, then mitigate the decision-induced downstream risks. To support this, we introduce WaymoQA, a dataset of 35,000 human-annotated question-answer pairs covering complex, high-risk driving scenarios. The dataset includes multiple-choice and open-ended formats across both image and video modalities. Experiments reveal that existing MLLMs underperform in safety-critical scenarios compared to normal scenes, but fine-tuning with WaymoQA significantly improves their reasoning ability, highlighting the effectiveness of our dataset in developing safer and more reasoning-capable driving agents. Our code and data are provided in https://github.com/sjyu001/WaymoQA
Authors: Rebeka Toth, Tamas Bisztray, Richard Dubniczky
Abstract: Phishing and spam emails remain a major cybersecurity threat, with attackers increasingly leveraging Large Language Models (LLMs) to craft highly deceptive content. This study presents a comprehensive email dataset containing phishing, spam, and legitimate messages, explicitly distinguishing between human- and LLM-generated content. Each email is annotated with its category, emotional appeal (e.g., urgency, fear, authority), and underlying motivation (e.g., link-following, credential theft, financial fraud). We benchmark multiple LLMs on their ability to identify these emotional and motivational cues and select the most reliable model to annotate the full dataset. To evaluate classification robustness, emails were also rephrased using several LLMs while preserving meaning and intent. A state-of-the-art LLM was then assessed on its performance across both original and rephrased emails using expert-labeled ground truth. The results highlight strong phishing detection capabilities but reveal persistent challenges in distinguishing spam from legitimate emails. Our dataset and evaluation framework contribute to improving AI-assisted email security systems. To support open science, all code, templates, and resources are available on our project site.
Authors: Mohamed Bouadi, Pratinav Seth, Aditya Tanna, Vinay Kumar Sankarapu
Abstract: Tabular data drive most real-world machine learning applications, yet building general-purpose models for them remains difficult. Mixed numeric and categorical fields, weak feature structure, and limited labeled data make scaling and generalization challenging. To this end, we introduce Orion-Bix, a tabular foundation model that combines biaxial attention with meta-learned in-context reasoning for few-shot tabular learning. Its encoder alternates standard, grouped, hierarchical, and relational attention, fusing their outputs through multi-CLS summarization to capture both local and global dependencies efficiently. A label-aware ICL head adapts on the fly and scales to large label spaces via hierarchical decision routing. Meta-trained on synthetically generated, structurally diverse tables with causal priors, Orion-Bix learns transferable inductive biases across heterogeneous data. Delivered as a scikit-learn compatible foundation model, it outperforms gradient-boosting baselines and remains competitive with state-of-the-art tabular foundation models on public benchmarks, showing that biaxial attention with episodic meta-training enables robust, few-shot-ready tabular learning. The model is publicly available at https://github.com/Lexsi-Labs/Orion-BiX .
Authors: Guangzhi Xiong, Zhenghao He, Bohan Liu, Sanchit Sinha, Aidong Zhang
Abstract: Retrieval-Augmented Generation (RAG) improves the factuality of large language models (LLMs) by grounding outputs in retrieved evidence, but faithfulness failures, where generations contradict or extend beyond the provided sources, remain a critical challenge. Existing hallucination detection methods for RAG often rely either on large-scale detector training, which requires substantial annotated data, or on querying external LLM judges, which leads to high inference costs. Although some approaches attempt to leverage internal representations of LLMs for hallucination detection, their accuracy remains limited. Motivated by recent advances in mechanistic interpretability, we employ sparse autoencoders (SAEs) to disentangle internal activations, successfully identifying features that are specifically triggered during RAG hallucinations. Building on a systematic pipeline of information-based feature selection and additive feature modeling, we introduce RAGLens, a lightweight hallucination detector that accurately flags unfaithful RAG outputs using LLM internal representations. RAGLens not only achieves superior detection performance compared to existing methods, but also provides interpretable rationales for its decisions, enabling effective post-hoc mitigation of unfaithful RAG. Finally, we justify our design choices and reveal new insights into the distribution of hallucination-related signals within LLMs. The code is available at https://github.com/Teddy-XiongGZ/RAGLens.
Authors: Tianjiao Yu, Xinzhuo Li, Yifan Shen, Yuanzhe Liu, Ismini Lourentzou
Abstract: Recent advances in large multimodal models suggest that explicit reasoning mechanisms play a critical role in improving model reliability, interpretability, and cross-modal alignment. While such reasoning-centric approaches have been proven effective in language and vision tasks, their extension to 3D remains underdeveloped. CoRe3D introduces a unified 3D understanding and generation reasoning framework that jointly operates over semantic and spatial abstractions, enabling high-level intent inferred from language to directly guide low-level 3D content formation. Central to this design is a spatially grounded reasoning representation that decomposes 3D latent space into localized regions, allowing the model to reason over geometry in a compositional and procedural manner. By tightly coupling semantic chain-of-thought inference with structured spatial reasoning, CoRe3D produces 3D outputs that exhibit strong local consistency and faithful alignment with linguistic descriptions.
Authors: Zishan Bai, Jing Luo, Ziyi Ni, Enze Ge, Jiacheng Shi, Yichao Zhang, Jiayi Gu, Zhimo Han, Riyang Bao, Junfeng Hao
Abstract: The proliferation of cloud-native architectures, characterized by microservices and dynamic orchestration, has rendered modern IT infrastructures exceedingly complex and volatile. This complexity generates overwhelming volumes of operational data, leading to critical bottlenecks in conventional systems: inefficient information processing, poor task coordination, and loss of contextual continuity during fault diagnosis and remediation. To address these challenges, we propose AOI (AI-Oriented Operations), a novel multi-agent collaborative framework that integrates three specialized agents with an LLM-based Context Compressor. Its core innovations include: (1) a dynamic task scheduling strategy that adaptively prioritizes operations based on real-time system states, (2) a three-layer memory architecture comprising Working, Episodic, and Semantic layers that optimizes context retention and retrieval. Extensive experiments on synthetic and real-world benchmarks show that AOI achieves 72.4\% context compression while preserving 92.8\% critical information, improves task success to 94.2\%, and reduces MTTR by 34.4\% over the best baseline. This work presents a paradigm shift towards scalable, adaptive, and context-aware autonomous operations, enabling robust management of next-generation IT infrastructures with minimal human intervention.
Authors: Scott Thornton
Abstract: AI coding assistants produce vulnerable code in 45\% of security-relevant scenarios~\cite{veracode2025}, yet no public training dataset teaches both traditional web security and AI/ML-specific defenses in a format suitable for instruction tuning. We present SecureCode, a production-grade dataset of 2,185 multi-turn security training examples spanning two domains: web application security (1,435 examples covering the OWASP Top 10 2021 across 11 languages and 9 frameworks, 100\% grounded in documented CVEs and security incidents) and AI/ML security (750 examples covering all 10 OWASP LLM Top 10 2025 categories across more than 40 frameworks, including LangChain, OpenAI, and Hugging Face). Every example follows a 4-turn conversational structure -- feature request; vulnerable and secure implementations with attack demonstrations; advanced probing; and defense-in-depth operational guidance -- designed for direct use in instruction tuning pipelines. Quality assurance combines automated structural validation with multi-agent review from seven specialist AI perspectives (more than 10{,}500 assessments) and an 8-phase remediation pipeline, producing a rubric-calibrated mean quality score of 93.8/100 ($\sigma = 0.93$) for the AI/ML component. Each example provides SIEM integration strategies, infrastructure hardening recommendations, and testing approaches using production frameworks. We release the unified dataset on Hugging Face with domain-specific loading configurations (web, aiml, default), alongside eight fine-tuned open-source models (3B--20B parameters, QLoRA), and an evaluation framework with four security-specific metrics. To our knowledge, SecureCode is the first public dataset that jointly provides OWASP Top 10 2021 web coverage and OWASP LLM Top 10 2025 AI/ML coverage in a unified conversational schema suitable for instruction tuning.
Authors: Chama Bensmail
Abstract: Machine learning models are primarily judged by predictive performance, especially in applied settings. Once a model reaches high accuracy, its explanation is often assumed to be correct and trustworthy. This assumption raises an overlooked question: when two models achieve high accuracy, do they rely on the same internal logic, or do they reach the same outcome via different and potentially competing mechanisms? We introduce EvoXplain, a diagnostic framework that measures the stability of model explanations across repeated training. Rather than analysing the explanation of a single trained model, EvoXplain treats explanations as samples drawn from the training and model selection pipeline itself, without aggregating predictions or constructing ensembles. It examines whether these samples form a single coherent explanatory basin or separate into multiple structured explanatory basins. We evaluate EvoXplain on the Adult Income and Breast Cancer datasets using deep neural networks and Logistic Regression. Although all models achieve high predictive accuracy, explanation stability differs across pipelines. Deep neural networks on Breast Cancer converge to a single explanatory basin, while the same architecture on Adult Income separates into distinct explanatory basins despite identical training conditions. Logistic Regression on Breast Cancer exhibits conditional multiplicity, where basin accessibility is controlled by regularisation configuration. EvoXplain does not attempt to select a correct explanation. Instead, it makes explanatory structure visible and quantifiable, revealing when single instance explanations obscure the existence of multiple admissible predictive mechanisms. More broadly, EvoXplain reframes interpretability as a property of the training pipeline under repeated instantiation, rather than of any single trained model.
Authors: Jos\'e Isidro, Filipe Cunha, Purifica\c{c}\~ao Silvano, Al\'ipio Jorge, Nuno Guimar\~aes, S\'ergio Nunes, Ricardo Campos
Abstract: Linear text segmentation is a long-standing problem in natural language processing (NLP), focused on dividing continuous text into coherent and semantically meaningful units. Despite its importance, the task remains challenging due to the complexity of defining topic boundaries, the variability in discourse structure, and the need to balance local coherence with global context. These difficulties hinder downstream applications such as summarization, information retrieval, and question answering. In this work, we introduce SegNSP, framing linear text segmentation as a next sentence prediction (NSP) task. Although NSP has largely been abandoned in modern pre-training, its explicit modeling of sentence-to-sentence continuity makes it a natural fit for detecting topic boundaries. We propose a label-agnostic NSP approach, which predicts whether the next sentence continues the current topic without requiring explicit topic labels, and enhance it with a segmentation-aware loss combined with harder negative sampling to better capture discourse continuity. Unlike recent proposals that leverage NSP alongside auxiliary topic classification, our approach avoids task-specific supervision. We evaluate our model against established baselines on two datasets, CitiLink-Minutes, for which we establish the first segmentation benchmark, and WikiSection. On CitiLink-Minutes, SegNSP achieves a B-$F_1$ of 0.79, closely aligning with human-annotated topic transitions, while on WikiSection it attains a B-F$_1$ of 0.65, outperforming the strongest reproducible baseline, TopSeg, by 0.17 absolute points. These results demonstrate competitive and robust performance, highlighting the effectiveness of modeling sentence-to-sentence continuity for improving segmentation quality and supporting downstream NLP applications.
Authors: Benjamin Rozonoyer, Chong You, Michael Boratko, Himanshu Jain, Nilesh Gupta, Srinadh Bhojanapalli, Andrew McCallum, Felix Yu
Abstract: The success of Large Language Models (LLMs) has motivated a shift toward generative approaches to retrieval and ranking, aiming to supersede classical Dual Encoders (DEs) and Cross Encoders (CEs). A prominent paradigm is pointwise Autoregressive Ranking (ARR), where an LLM generates document identifiers (docIDs) token-by-token to enable ranking via beam search. ARR offers the promise of superior expressivity compared to DEs while avoiding the prohibitive computational cost of CEs. However, a formal theoretical foundation for this expressive power has been missing. Moreover, the standard next-token prediction loss is rank-agnostic and inappropriate for finetuning an LLM for ranking tasks. In this paper, we first prove that the expressive capacity of ARR is strictly superior to DEs. While a DE requires an embedding dimension that grows linearly with corpus size to achieve arbitrary rankings, ARR can solve it with a constant hidden dimension. We then propose SToICaL (Simple Token-Item Calibrated Loss), a generalized rank-aware training loss for LLM finetuning. By using item-level reweighting and prefix-tree marginalization, we distribute probability mass over valid docID tokens based on their ground-truth relevance. Experiments on WordNet and ESCI datasets verify that our loss suppresses invalid docID generations and significantly improves ranking metrics beyond top-1 retrieval.
Authors: Sung Jun Cheon, Jaekyung Cho, Seongho Choi, Hyunjun Eun, Seokhwan Jo, Jaehyun Jun, Minsoo Kang, Jin Kim, Jiwon Kim, Minsang Kim, Seungsik Kim, Sungwan Kim, Tae Yoon Kim, Youngrang Kim, Hyeongmun Lee, Sangyeol Lee, Sungeun Lee, Youngsoon Lee, Yujin Lee, Seongmin Ok, Chanyong Park, Hyewoong Park, Junyoung Park, Hyunho Yang, Subin Yi, Dhammiko Arya, Soohyun Bae, Dongyeon Cho, Seungmo Cho, Sangho Choi, Yongseok Choi, Gyoungeun Han, Yong-jin Han, Seokyoung Hong, Hyeon Hwang, Wonbeom Jang, Minjeong Ju, Wonjin Jung, Keummin Ka, Sungil Kang, Dongnam Kim, Jonghwi Kim, Joonghoon Kim, SaeRom Kim, Sangjin Kim, Seongwon Kim, Youngjin Kim, Seojin Lee, Sunwoo Lee, Taehoon Lee, Chanwoo Park, Sohee Park, Sooyeon Park, Yohan Ra, Sereimony Sek, Seungyeon Seo, Gun Song, Sanghoon Woo, Janghan Yoon, Sungbin Yoon
Abstract: We introduce A.X K1, a 519B-parameter Mixture-of-Experts (MoE) language model trained from scratch. Our design leverages scaling laws to optimize training configurations and vocabulary size under fixed computational budgets. A.X K1 is pre-trained on a corpus of approximately 10T tokens, curated by a multi-stage data processing pipeline. Designed to bridge the gap between reasoning capability and inference efficiency, A.X K1 supports explicitly controllable reasoning to facilitate scalable deployment across diverse real-world scenarios. We propose a simple yet effective Think-Fusion training recipe, enabling user-controlled switching between thinking and non-thinking modes within a single unified model. Extensive evaluations demonstrate that A.X K1 achieves performance competitive with leading open-source models, while establishing a distinctive advantage in Korean-language benchmarks.
Authors: Jeremiah Coholich, Justin Wit, Robert Azarcon, Zsolt Kira
Abstract: Vision-based policies for robot manipulation have achieved significant recent success, but are still brittle to distribution shifts such as camera viewpoint variations. Robot demonstration data is scarce and often lacks appropriate variation in camera viewpoints. Simulation offers a way to collect robot demonstrations at scale with comprehensive coverage of different viewpoints, but presents a visual sim2real challenge. To bridge this gap, we propose MANGO -- an unpaired image translation method with a novel segmentation-conditioned InfoNCE loss, a highly-regularized discriminator design, and a modified PatchNCE loss. We find that these elements are crucial for maintaining viewpoint consistency during sim2real translation. When training MANGO, we only require a small amount of fixed-camera data from the real world, but show that our method can generate diverse unseen viewpoints by translating simulated observations. In this domain, MANGO outperforms all other image translation methods we tested. Imitation-learning policies trained on data augmented by MANGO are able to achieve success rates as high as 60% on views that the non-augmented policy fails completely on.
Authors: Jing-Yi Zeng, Guan-Hua Huang
Abstract: This study investigates how to efficiently build a domain-specialized large language model (LLM) for statistics using the lightweight LLaMA-3.2-3B family as the foundation model (FM). We systematically compare three multi-stage training pipelines--starting from a base FM with no instruction-following capability, a base FM augmented with post-hoc instruction tuning, and an instruction-tuned FM with strong general reasoning abilities--across continual pretraining, supervised fine-tuning (SFT), reinforcement learning from human feedback (RLHF) preference alignment, and downstream task fine-tuning (DTFT). Results show that pipelines beginning with a base FM fail to develop meaningful statistical reasoning, even after extensive instruction tuning, SFT, or RLHF alignment. In contrast, starting from LLaMA-3.2-3B-Instruct enables effective domain specialization. A comprehensive evaluation of SFT variants reveals clear trade-offs between domain expertise and general reasoning ability. We further demonstrate that direct preference optimization provides stable and effective RLHF preference alignment. Finally, we show that DTFT must be performed with extremely low intensity to avoid catastrophic forgetting in highly optimized models. The final model, StatLLaMA, achieves strong and balanced performance on benchmarks of mathematical reasoning, common-sense reasoning, and statistical expertise, offering a practical blueprint for developing resource-efficient statistical LLMs. The code is available at https://github.com/HuangDLab/StatLLaMA.
Authors: J\'anos Kram\'ar, Joshua Engels, Zheng Wang, Bilal Chughtai, Rohin Shah, Neel Nanda, Arthur Conmy
Abstract: Frontier language model capabilities are improving rapidly. We thus need stronger mitigations against bad actors misusing increasingly powerful systems. Prior work has shown that activation probes may be a promising misuse mitigation technique, but we identify a key remaining challenge: probes fail to generalize under important production distribution shifts. In particular, we find that the shift from short-context to long-context inputs is difficult for existing probe architectures. We propose several new probe architectures that handle this long-context distribution shift. We evaluate these probes in the cyber-offensive domain, testing their robustness against various production-relevant distribution shifts, including multi-turn conversations, long context prompts, and adaptive red teaming. Our results demonstrate that while our novel architectures address context length, a combination of architecture choice and training on diverse distributions is required for broad generalization. Additionally, we show that pairing probes with prompted classifiers achieves optimal accuracy at a low cost due to the computational efficiency of probes. These findings have informed the successful deployment of misuse mitigation probes in user-facing instances of Gemini, Google's frontier language model. Finally, we find early positive results using AlphaEvolve to automate improvements in both probe architecture search and adaptive red teaming, showing that automating some AI safety research is already possible.
Authors: Liz Li, Wei Zhu
Abstract: While Retrieval-Augmented Generation (RAG) has been swiftly adopted in scientific and clinical QA systems, a comprehensive evaluation benchmark in the medical domain is lacking. To address this gap, we introduce the Medical Retrieval-Augmented Generation (MRAG) benchmark, covering various tasks in English and Chinese languages, and building a corpus with Wikipedia and Pubmed. Additionally, we develop the MRAG-Toolkit, facilitating systematic exploration of different RAG components. Our experiments reveal that: (a) RAG enhances LLM reliability across MRAG tasks. (b) the performance of RAG systems is influenced by retrieval approaches, model sizes, and prompting strategies. (c) While RAG improves usefulness and reasoning quality, LLM responses may become slightly less readable for long-form questions. We will release the MRAG-Bench's dataset and toolkit with CCBY-4.0 license upon acceptance, to facilitate applications from both academia and industry.
Authors: Jamie Duell, Xiuyi Fan
Abstract: Counterfactual explanations (CEs) offer interpretable insights into machine learning predictions by answering ``what if?" questions. However, in real-world settings where models are frequently updated, existing counterfactual explanations can quickly become invalid or unreliable. In this paper, we introduce Probabilistically Safe CEs (PSCE), a method for generating counterfactual explanations that are $\delta$-safe, to ensure high predictive confidence, and $\epsilon$-robust to ensure low predictive variance. Based on Bayesian principles, PSCE provides formal probabilistic guarantees for CEs under model changes which are adhered to in what we refer to as the $\langle \delta, \epsilon \rangle$-set. Uncertainty-aware constraints are integrated into our optimization framework and we validate our method empirically across diverse datasets. We compare our approach against state-of-the-art Bayesian CE methods, where PSCE produces counterfactual explanations that are not only more plausible and discriminative, but also provably robust under model change.
Authors: Junyi Zou
Abstract: Large language models fine-tuned via a two-stage pipeline (domain adaptation followed by instruction alignment) can exhibit non-trivial interference after adapter merging, including the re-emergence of explicit reasoning traces under strict decoding. We study this phenomenon in medical LLM settings using lightweight, reproducible measurements of trace leakage and instruction-following behavior. Beyond marker-based proxies, we introduce a marker-forbidden, answer-only evaluation and define a correctness-based direction that does not rely on surface markers; a rank-1 logit-space intervention along this direction modulates decision distributions and improves multiple-choice accuracy beyond random-direction controls at sufficiently large intervention strength. We further provide layer-wise geometric evidence that domain and instruction adapters induce partially misaligned update directions, and present a proof-of-concept geometry-aware merge that can reduce leakage and/or improve accuracy in a toy setting. Our results characterize boundary conditions of trace leakage and provide practical diagnostics and interventions for safer adapter merging.
Authors: Yingxiao Huo, Satya Prakash Dash, Radu Stoican, Samuel Kaski, Mingfei Sun
Abstract: Natural gradients have long been studied in deep reinforcement learning due to their fast convergence properties and covariant weight updates. However, computing natural gradients requires inversion of the Fisher Information Matrix (FIM) at each iteration, which is computationally prohibitive in nature. In this paper, we present an efficient and scalable natural policy optimization technique that leverages a rank-1 approximation to full inverse-FIM. We theoretically show that under certain conditions, a rank-1 approximation to inverse-FIM converges faster than policy gradients and, under some conditions, enjoys the same sample complexity as stochastic policy gradient methods. We benchmark our method on a diverse set of environments and show that it achieves superior performance to standard actor-critic and trust-region baselines.
Authors: Hong Liu, Jiaqi Zhang, Chao Wang, Xing Hu, Linkun Lyu, Jiaqi Sun, Xurui Yang, Bo Wang, Fengcun Li, Yulei Qian, Lingtong Si, Yerui Sun, Rumei Li, Peng Pei, Yuchen Xie, Xunliang Cai
Abstract: While Mixture-of-Experts (MoE) architectures have become the standard for sparsity scaling in large language models, they increasingly face diminishing returns and system-level bottlenecks. In this work, we explore embedding scaling as a potent, orthogonal dimension for scaling sparsity. Through a comprehensive analysis and experiments, we identify specific regimes where embedding scaling achieves a superior Pareto frontier compared to expert scaling. We systematically characterize the critical architectural factors governing this efficacy -- ranging from parameter budgeting to the interplay with model width and depth. Moreover, by integrating tailored system optimizations and speculative decoding, we effectively convert this sparsity into tangible inference speedups. Guided by these insights, we introduce LongCat-Flash-Lite, a 68.5B parameter model with ~3B activated trained from scratch. Despite allocating over 30B parameters to embeddings, LongCat-Flash-Lite not only surpasses parameter-equivalent MoE baselines but also exhibits exceptional competitiveness against existing models of comparable scale, particularly in agentic and coding domains.
Authors: Alexander Loth, Martin Kappes, Marc-Oliver Pahl
Abstract: Generative AI and misinformation research has evolved since our 2024 survey. This paper presents an updated perspective, transitioning from literature review to practical countermeasures. We report on changes in the threat landscape, including improved AI-generated content through Large Language Models (LLMs) and multimodal systems. Central to this work are our practical contributions: JudgeGPT, a platform for evaluating human perception of AI-generated news, and RogueGPT, a controlled stimulus generation engine for research. Together, these tools form an experimental pipeline for studying how humans perceive and detect AI-generated misinformation. Our findings show that detection capabilities have improved, but the competition between generation and detection continues. We discuss mitigation strategies including LLM-based detection, inoculation approaches, and the dual-use nature of generative AI. This work contributes to research addressing the adverse impacts of AI on information quality.
Authors: Chanwook Park, Brian Kim, Jiachen Guo, Wing Kam Liu
Abstract: Neural networks and machine learning models for uncertainty quantification suffer from limited scalability and poor reliability compared to their deterministic counterparts. In industry-scale active learning settings, where generating a single high-fidelity simulation may require days or weeks of computation and produce data volumes on the order of gigabytes, they quickly become impractical. This paper proposes a scalable and reliable Bayesian surrogate model, termed the Bayesian Interpolating Neural Network (B-INN). The B-INN combines high-order interpolation theory with tensor decomposition and alternating direction algorithm to enable effective dimensionality reduction without compromising predictive accuracy. We theoretically show that the function space of a B-INN is a subset of that of Gaussian processes, while its Bayesian inference exhibits linear complexity, $\mathcal{O}(N)$, with respect to the number of training samples. Numerical experiments demonstrate that B-INNs can be from 20 times to 10,000 times faster with a robust uncertainty estimation compared to Bayesian neural networks and Gaussian processes. These capabilities make B-INN a practical foundation for uncertainty-driven active learning in large-scale industrial simulations, where computational efficiency and robust uncertainty calibration are paramount.
Authors: Afrozah Nadeem, Agrima Seth, Mehwish Nasim, Usman Naseem
Abstract: Large Language Models (LLMs) increasingly shape global discourse, making fairness and ideological neutrality essential for responsible AI deployment. Despite growing attention to political bias in LLMs, prior work largely focuses on high-resource, Western languages or narrow multilingual settings, leaving cross-lingual consistency and safe post-hoc mitigation underexplored. To address this gap, we present a large-scale multilingual evaluation of political bias spanning 50 countries and 33 languages. We introduce a complementary post-hoc mitigation framework, Cross-Lingual Alignment Steering (CLAS), designed to augment existing steering methods by aligning ideological representations across languages and dynamically regulating intervention strength. This method aligns latent ideological representations induced by political prompts into a shared ideological subspace, ensuring cross lingual consistency, with the adaptive mechanism prevents over correction and preserves coherence. Experiments demonstrate substantial bias reduction along both economic and social axes with minimal degradation in response quality. The proposed framework establishes a scalable and interpretable paradigm for fairness-aware multilingual LLM governance, balancing ideological neutrality with linguistic and cultural diversity.
Authors: Keishu Utimula
Abstract: Large Language Models (LLMs), trained solely on massive text data, have achieved high performance on the Winograd Schema Challenge (WSC), a benchmark proposed to measure commonsense knowledge and reasoning abilities about the real world. This suggests that the language produced by humanity describes a significant portion of the world with considerable nuance. In this study, we attempt to harness the high expressive power of language within cellular automata. Specifically, we express cell states and rules in natural language and delegate their updates to an LLM. Through this approach, cellular automata can transcend the constraints of merely numerical states and fixed rules, providing us with a richer platform for simulation. Here, we propose LOGOS-CA (Language Oriented Grid Of Statements - Cellular Automaton) as a natural framework to achieve this and examine its capabilities. We confirmed that LOGOS-CA successfully performs simple forest fire simulations and also serves as an intriguing subject for investigation from an Artificial Life (ALife) perspective. In this paper, we report the results of these experiments and discuss directions for future research using LOGOS-CA.
Authors: Hang Wu, Yujun Cai, Zehao Li, Haonan Ge, Bowen Sun, Junsong Yuan, Yiwei Wang
Abstract: Understanding camera dynamics is a fundamental pillar of video spatial intelligence. However, existing multimodal models predominantly treat this task as a black-box classification, often confusing physically distinct motions by relying on superficial visual patterns rather than geometric cues. We present CamReasoner, a framework that reformulates camera movement understanding as a structured inference process to bridge the gap between perception and cinematic logic. Our approach centers on the Observation-Thinking-Answer (O-T-A) paradigm, which compels the model to decode spatio-temporal cues such as trajectories and view frustums within an explicit reasoning block. To instill this capability, we construct a Large-scale Inference Trajectory Suite comprising 18k SFT reasoning chains and 38k RL feedback samples. Notably, we are the first to employ RL for logical alignment in this domain, ensuring motion inferences are grounded in physical geometry rather than contextual guesswork. By applying Reinforcement Learning to the Observation-Think-Answer (O-T-A) reasoning paradigm, CamReasoner effectively suppresses hallucinations and achieves state-of-the-art performance across multiple benchmarks.
Authors: Shreshth Saini, Avinab Saha, Balu Adsumilli, Neil Birkbeck, Yilin Wang, Alan C. Bovik
Abstract: Masked Diffusion Models (MDMs) have emerged as a promising non-autoregressive paradigm for generative tasks, offering parallel decoding and bidirectional context utilization. However, current sampling methods rely on simple confidence-based heuristics that ignore the long-term impact of local decisions, leading to trajectory lock-in where early hallucinations cascade into global incoherence. While search-based methods mitigate this, they incur prohibitive computational costs ($O(K)$ forward passes per step). In this work, we propose Backward-on-Entropy (BoE) Steering, a gradient-guided inference framework that approximates infinite-horizon lookahead via a single backward pass. We formally derive the Token Influence Score (TIS) from a first-order expansion of the trajectory cost functional, proving that the gradient of future entropy with respect to input embeddings serves as an optimal control signal for minimizing uncertainty. To ensure scalability, we introduce \texttt{ActiveQueryAttention}, a sparse adjoint primitive that exploits the structure of the masking objective to reduce backward pass complexity. BoE achieves a superior Pareto frontier for inference-time scaling compared to existing unmasking methods, demonstrating that gradient-guided steering offers a mathematically principled and efficient path to robust non-autoregressive generation. We will release the code.
Authors: Aditya Kumar Purohit, Hendrik Heuer
Abstract: Large Language Models (LLMs) are increasingly used for mental health support, yet little is known about how people with mental health challenges engage with them, how they evaluate their usefulness, and what design opportunities they envision. We conducted 20 semi-structured interviews with people in the UK who live with mental health conditions and have used LLMs for mental health support. Through reflexive thematic analysis, we found that participants engaged with LLMs in conditional and situational ways: for immediacy, the desire for non-judgement, self-paced disclosure, cognitive reframing, and relational engagement. Simultaneously, participants articulated clear boundaries informed by prior therapeutic experience: LLMs were effective for mild-to-moderate distress but inadequate for crises, trauma, and complex social-emotional situations. We contribute empirical insights into the lived use of LLMs for mental health, highlight boundary-setting as central to their safe role, and propose design and governance directions for embedding them responsibly within care ecosystem.
Authors: Dana Feng, Bhada Yun, April Yi Wang
Abstract: Juniors enter as AI-natives, seniors adapted mid-career. AI is not just changing how engineers code-it is reshaping who holds agency across work and professional growth. We contribute junior-senior accounts on their usage of agentic AI through a three-phase mixed-methods study: ACTA combined with a Delphi process with 5 seniors, an AI-assisted debugging task with 10 juniors, and blind reviews of junior prompt histories by 5 more seniors. We found that agency in software engineering is primarily constrained by organizational policies rather than individual preferences, with experienced developers maintaining control through detailed delegation while novices struggle between over-reliance and cautious avoidance. Seniors leverage pre-AI foundational instincts to steer modern tools and possess valuable perspectives for mentoring juniors in their early AI-encouraged career development. From synthesis of results, we suggest three practices that focus on preserving agency in software engineering for coding, learning, and mentorship, especially as AI grows increasingly autonomous.
Authors: Lukas Kuhn, Giuseppe Serra, Florian Buettner
Abstract: Vision-language models have transformed multimodal representation learning, yet dominant contrastive approaches like CLIP require large batch sizes, careful negative sampling, and extensive hyperparameter tuning. We introduce NOVA, a NOn-contrastive Vision-language Alignment framework based on joint embedding prediction with distributional regularization. NOVA aligns visual representations to a frozen, domain-specific text encoder by predicting text embeddings from augmented image views, while enforcing an isotropic Gaussian structure via Sketched Isotropic Gaussian Regularization (SIGReg). This eliminates the need for negative sampling, momentum encoders, or stop-gradients, reducing the training objective to a single hyperparameter. We evaluate NOVA on zeroshot chest X-ray classification using ClinicalBERT as the text encoder and Vision Transformers trained from scratch on MIMIC-CXR. On zero-shot classification across three benchmark datasets, NOVA outperforms multiple standard baselines while exhibiting substantially more consistent training runs. Our results demonstrate that non-contrastive vision-language pretraining offers a simpler, more stable, and more effective alternative to contrastive methods.
Authors: Nima Shoghi, Yuxuan Liu, Yuning Shen, Rob Brekelmans, Pan Li, Quanquan Gu
Abstract: Molecular dynamics (MD) simulations remain the gold standard for studying protein dynamics, but their computational cost limits access to biologically relevant timescales. Recent generative models have shown promise in accelerating simulations, yet they struggle with long-horizon generation due to architectural constraints, error accumulation, and inadequate modeling of spatio-temporal dynamics. We present STAR-MD (Spatio-Temporal Autoregressive Rollout for Molecular Dynamics), a scalable SE(3)-equivariant diffusion model that generates physically plausible protein trajectories over microsecond timescales. Our key innovation is a causal diffusion transformer with joint spatio-temporal attention that efficiently captures complex space-time dependencies while avoiding the memory bottlenecks of existing methods. On the standard ATLAS benchmark, STAR-MD achieves state-of-the-art performance across all metrics--substantially improving conformational coverage, structural validity, and dynamic fidelity compared to previous methods. STAR-MD successfully extrapolates to generate stable microsecond-scale trajectories where baseline methods fail catastrophically, maintaining high structural quality throughout the extended rollout. Our comprehensive evaluation reveals severe limitations in current models for long-horizon generation, while demonstrating that STAR-MD's joint spatio-temporal modeling enables robust dynamics simulation at biologically relevant timescales, paving the way for accelerated exploration of protein function.
Authors: Ron Shprints, Peter Holderrieth, Juno Nam, Rafael G\'omez-Bombarelli, Tommi Jaakkola
Abstract: Transition states (TSs) are central to understanding and quantitatively predicting chemical reactivity and reaction mechanisms. Although traditional TS generation methods are computationally expensive, recent generative modeling approaches have enabled chemically meaningful TS prediction for relatively small molecules. However, these methods fail to generalize to practically relevant reaction substrates because of distribution shifts induced by increasing molecular sizes. Furthermore, TS geometries for larger molecules are not available at scale, making it infeasible to train generative models from scratch on such molecules. To address these challenges, we introduce FragmentFlow: a divide-and-conquer approach that trains a generative model to predict TS geometries for the reactive core atoms, which define the reaction mechanism. The full TS structure is then reconstructed by re-attaching substituent fragments to the predicted core. By operating on reactive cores, whose size and composition remain relatively invariant across molecular contexts, FragmentFlow mitigates distribution shifts in generative modeling. Evaluated on a new curated dataset of reactions involving reactants with up to 33 heavy atoms, FragmentFlow correctly identifies 90% of TSs while requiring 30% fewer saddle-point optimization steps than classical initialization schemes. These results point toward scalable TS generation for high-throughput reactivity studies.
Authors: Bizhe Bai, Xinyue Wang, Peng Ye, Tao Chen
Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) improves LLM reasoning, yet growing evidence indicates an exploration ceiling: it often reweights existing solution traces rather than discovering new strategies, limiting gains under large sampling budgets (e.g., pass-at-256). We address this limitation with PSN-RLVR, which perturbs policy parameters before rollout generation to induce temporally consistent, trajectory-level exploration that better preserves long-horizon chain-of-thought coherence than action-space noise. To mitigate the resulting sampling-update mismatch, we incorporate truncated importance sampling (TIS). To avoid expensive KL-based adaptive noise control, we propose a computationally efficient real-time adaptive noise scheduler driven by a lightweight surrogate that combines semantic diversity with normalized self-certainty. Instantiated on GRPO, a widely used RLVR method, PSN-GRPO consistently expands the effective reasoning capability boundary across multiple mathematical reasoning benchmarks and model families, yielding higher pass-at-k under large sampling budgets and outperforming prior exploration-oriented RLVR methods (e.g., Pass-at-k-style training) while remaining orthogonal and thus composable for additional gains.
Authors: Zihao Jing, Qiuhao Zeng, Ruiyi Fang, Yan Sun, Boyu Wang, Pingzhao Hu
Abstract: Molecular understanding is central to advancing areas such as scientific discovery, yet Large Language Models (LLMs) struggle to understand molecular graphs effectively. Existing graph-LLM bridges often adapt the Q-Former-style connector with fixed-length static tokens, which is originally designed for vision tasks. These designs overlook stereochemistry and substructural context and typically require costly LLM-backbone fine-tuning, limiting efficiency and generalization. We introduce EDT-Former, an Entropy-guided Dynamic Token Transformer that generates tokens aligned with informative molecular patches, thereby preserving both local and global structural features for molecular graph understanding. Beyond prior approaches, EDT-Former enables alignment between frozen graph encoders and LLMs without tuning the LLM backbone (excluding the embedding layer), resulting in computationally efficient finetuning, and achieves stateof-the-art results on MoleculeQA, Molecule-oriented Mol-Instructions, and property prediction benchmarks (TDC, MoleculeNet), underscoring its effectiveness for scalable and generalizable multimodal molecular understanding
Authors: Yiran Qiao, Jing Chen, Xiang Ao, Qiwei Zhong, Yang Liu, Qing He
Abstract: Live streaming has become a cornerstone of today's internet, enabling massive real-time social interactions. However, it faces severe risks arising from sparse, coordinated malicious behaviors among multiple participants, which are often concealed within normal activities and challenging to detect timely and accurately. In this work, we provide a pioneering study on risk assessment in live streaming rooms, characterized by weak supervision where only room-level labels are available. We formulate the task as a Multiple Instance Learning (MIL) problem, treating each room as a bag and defining structured user-timeslot capsules as instances. These capsules represent subsequences of user actions within specific time windows, encapsulating localized behavioral patterns. Based on this formulation, we propose AC-MIL, an Action-aware Capsule MIL framework that models both individual behaviors and group-level coordination patterns. AC-MIL captures multi-granular semantics and behavioral cues through a serial and parallel architecture that jointly encodes temporal dynamics and cross-user dependencies. These signals are integrated for robust room-level risk prediction, while also offering interpretable evidence at the behavior segment level. Extensive experiments on large-scale industrial datasets from Douyin demonstrate that AC-MIL significantly outperforms MIL and sequential baselines, establishing new state-of-the-art performance in room-level risk assessment for live streaming. Moreover, AC-MIL provides capsule-level interpretability, enabling identification of risky behavior segments as actionable evidence for intervention. The project page is available at: https://qiaoyran.github.io/AC-MIL/.
Authors: Tao Yu, Minghui Zhang, Zhiqing Cui, Hao Wang, Zhongtian Luo, Shenghua Chai, Junhao Gong, Yuzhao Peng, Yuxuan Zhou, Yujia Yang, Zhenghao Zhang, Haopeng Jin, Xinming Wang, Yufei Xiong, Jiabing Yang, Jiahao Yuan, Hanqing Wang, Hongzhu Yi, Yan Huang, Liang Wang
Abstract: Transforming scientific papers into multimodal presentation content is essential for research dissemination but remains labor intensive. Existing automated solutions typically treat each format as an isolated downstream task, leading to redundant processing and semantic inconsistency. We introduce PaperX, a unified framework that models academic presentation generation as a structural transformation and rendering process. Central to our approach is the Scholar DAG, an intermediate representation that decouples the paper's logical structure from its final presentation syntax. By applying adaptive graph traversal strategies, PaperX generates diverse, high quality outputs from a single source. Comprehensive evaluations demonstrate that our framework achieves the state of the art performance in content fidelity and aesthetic quality while significantly improving cost efficiency compared to specialized single task agents.
Authors: Chaoqi Liu, Xiaoshen Han, Jiawei Gao, Yue Zhao, Haonan Chen, Yilun Du
Abstract: Autoregressive policies offer a compelling foundation for scalable robot learning by enabling discrete abstraction, token-level reasoning, and flexible inference. However, applying autoregressive modeling to continuous robot actions requires an effective action tokenization scheme. Existing approaches either rely on analytical discretization methods that produce prohibitively long token sequences, or learned latent tokenizers that lack structure, limiting their compatibility with next-token prediction. In this work, we identify three desiderata for action tokenization - high compression, total decodability, and a left-to-right causally ordered token space - and introduce Ordered Action Tokenization (OAT), a learned action tokenizer that satisfies all three. OAT discretizes action chunks into an ordered sequence of tokens using transformer with registers, finite scalar quantization, and ordering-inducing training mechanisms. The resulting token space aligns naturally with autoregressive generation and enables prefix-based detokenization, yielding an anytime trade-off between inference cost and action fidelity. Across more than 20 tasks spanning four simulation benchmarks and real-world settings, autoregressive policies equipped with OAT consistently outperform prior tokenization schemes and diffusion-based baselines, while offering significantly greater flexibility at inference time.
Authors: Sai Sindhur Malleni, Ra\'ul Sevilla, Aleksei Vasilevskii, Jos\'e Castillo Lema, Andr\'e Bauer
Abstract: As Generative AI (GenAI), particularly inference, rapidly emerges as a dominant workload category, the Kubernetes ecosystem is proactively evolving to natively support its unique demands. This industry paper demonstrates how emerging Kubernetes-native projects can be combined to deliver the benefits of container orchestration, such as scalability and resource efficiency, to complex AI workflows. We implement and evaluate an illustrative, multi-stage use case consisting of automatic speech recognition and summarization. First, we address batch inference by using Kueue to manage jobs that transcribe audio files with Whisper models and Dynamic Accelerator Slicer (DAS) to increase parallel job execution. Second, we address a discrete online inference scenario by feeding the transcripts to a Large Language Model for summarization hosted using llm-d, a novel solution utilizing the recent developments around the Kubernetes Gateway API Inference Extension (GAIE) for optimized routing of inference requests. Our findings illustrate that these complementary components (Kueue, DAS, and GAIE) form a cohesive, high-performance platform, proving Kubernetes' capability to serve as a unified foundation for demanding GenAI workloads: Kueue reduced total makespan by up to 15%; DAS shortened mean job completion time by 36\%; and GAIE working in conjunction with llm-d improved tail Time to First Token latency by up to 90% even under high loads.
Authors: Eryue Xu, Tianshi Li
Abstract: Managing one's digital footprint is overwhelming, as it spans multiple platforms and involves countless context-dependent decisions. Recent advances in agentic AI offer ways forward by enabling holistic, contextual privacy-enhancing solutions. Building on this potential, we adopted a ''human-as-the-unit'' perspective and investigated users' cross-context privacy challenges through 12 semi-structured interviews. Results reveal that people rely on ad hoc manual strategies while lacking comprehensive privacy controls, highlighting nine privacy-management challenges across applications, temporal contexts, and relationships. To explore solutions, we generated nine AI agent concepts and evaluated them via a speed-dating survey with 116 US participants. The three highest-ranked concepts were all post-sharing management tools with half or full agent autonomy, with users expressing greater trust in AI accuracy than in their own efforts. Our findings highlight a promising design space where users see AI agents bridging the fragments in privacy management, particularly through automated, comprehensive post-sharing remediation of users' digital footprints.
Authors: Xiandong Zou, Jianshu Li, Jing Huang, Pan Zhou
Abstract: Speculative decoding accelerates inference for (M)LLMs, yet a training-decoding discrepancy persists: while existing methods optimize single greedy trajectories, decoding involves verifying and ranking multiple sampled draft paths. We propose Variational Speculative Decoding (VSD), formulating draft training as variational inference over latent proposals (draft paths). VSD maximizes the marginal probability of target-model acceptance, yielding an ELBO that promotes high-quality latent proposals while minimizing divergence from the target distribution. To enhance quality and reduce variance, we incorporate a path-level utility and optimize via an Expectation-Maximization procedure. The E-step draws MCMC samples from an oracle-filtered posterior, while the M-step maximizes weighted likelihood using Adaptive Rejection Weighting (ARW) and Confidence-Aware Regularization (CAR). Theoretical analysis confirms that VSD increases expected acceptance length and speedup. Extensive experiments across LLMs and MLLMs show that VSD achieves up to a 9.6% speedup over EAGLE-3 and 7.9% over ViSpec, significantly improving decoding efficiency.
Authors: Ehud Shapiro
Abstract: Grassroots Logic Programs (GLP) is a concurrent logic programming language with variables partitioned into paired \emph{readers} and \emph{writers}, conjuring both linear logic and futures/promises: an assignment is produced at most once via the sole occurrence of a writer (promise) and consumed at most once via the sole occurrence of its paired reader (future), and may contain additional readers and/or writers, enabling the concise expression of rich multidirectional communication modalities. GLP was designed as a language for grassroots platforms -- distributed systems with multiple instances that can operate independently of each other and of any global resource, and can coalesce into ever larger instances -- with its target architecture being smartphones communicating peer-to-peer. The operational semantics of Concurrent (single-agent) GLP and of multiagent GLP (maGLP) were defined via transition systems/multiagent transition systems, respectively. Here, we describe the mathematics developed to facilitate the workstation- and smartphone-based implementations of GLP by AI in Dart. We developed dGLP -- implementation-ready deterministic operational semantics for single-agent GLP -- and proved it correct with respect to the Concurrent GLP operational semantics; dGLP was used by AI as a formal spec, from which it developed a workstation-based implementation of GLP. We developed madGLP -- an implementation-ready multiagent operational semantics for maGLP -- and proved it correct with respect to the maGLP operational semantics; madGLP is deterministic at the agent level (not at the system level due to communication asynchrony), and is being used by AI as a formal spec from which it develops a smartphone-based implementation of maGLP.
Authors: Krzysztof Wr\'obel, Jan Maria Kowalski, Jerzy Surma, Igor Ciuciura, Maciej Szyma\'nski
Abstract: As Large Language Models (LLMs) become increasingly deployed in Polish language applications, the need for efficient and accurate content safety classifiers has become paramount. We present Bielik Guard, a family of compact Polish language safety classifiers comprising two model variants: a 0.1B parameter model based on MMLW-RoBERTa-base and a 0.5B parameter model based on PKOBP/polish-roberta-8k. Fine-tuned on a community-annotated dataset of 6,885 Polish texts, these models classify content across five safety categories: Hate/Aggression, Vulgarities, Sexual Content, Crime, and Self-Harm. Our evaluation demonstrates that both models achieve strong performance on multiple benchmarks. The 0.5B variant offers the best overall discrimination capability with F1 scores of 0.791 (micro) and 0.785 (macro) on the test set, while the 0.1B variant demonstrates exceptional efficiency. Notably, Bielik Guard 0.1B v1.1 achieves superior precision (77.65%) and very low false positive rate (0.63%) on real user prompts, outperforming HerBERT-PL-Guard (31.55% precision, 4.70% FPR) despite identical model size. The models are publicly available and designed to provide appropriate responses rather than simple content blocking, particularly for sensitive categories like self-harm.
Authors: Yixuan Ye, Xuanyu Lu, Yuxin Jiang, Yuchao Gu, Rui Zhao, Qiwei Liang, Jiachun Pan, Fengda Zhang, Weijia Wu, Alex Jinpeng Wang
Abstract: World models aim to understand, remember, and predict dynamic visual environments, yet a unified benchmark for evaluating their fundamental abilities remains lacking. To address this gap, we introduce MIND, the first open-domain closed-loop revisited benchmark for evaluating Memory consIstency and action coNtrol in worlD models. MIND contains 250 high-quality videos at 1080p and 24 FPS, including 100 (first-person) + 100 (third-person) video clips under a shared action space and 25 + 25 clips across varied action spaces covering eight diverse scenes. We design an efficient evaluation framework to measure two core abilities: memory consistency and action control, capturing temporal stability and contextual coherence across viewpoints. Furthermore, we design various action spaces, including different character movement speeds and camera rotation angles, to evaluate the action generalization capability across different action spaces under shared scenes. To facilitate future performance benchmarking on MIND, we introduce MIND-World, a novel interactive Video-to-World baseline. Extensive experiments demonstrate the completeness of MIND and reveal key challenges in current world models, including the difficulty of maintaining long-term memory consistency and generalizing across action spaces. Code: https://github.com/CSU-JPG/MIND.
Authors: Paul Saegert, Ullrich K\"othe
Abstract: Symbolic regression (SR) aims to discover interpretable analytical expressions that accurately describe observed data. Amortized SR promises to be much more efficient than the predominant genetic programming SR methods, but currently struggles to scale to realistic scientific complexity. We find that a key obstacle is the lack of a fast reduction of equivalent expressions to a concise normalized form. Amortized SR has addressed this by general-purpose Computer Algebra Systems (CAS) like SymPy, but the high computational cost severely limits training and inference speed. We propose SimpliPy, a rule-based simplification engine achieving a 100-fold speed-up over SymPy at comparable quality. This enables substantial improvements in amortized SR, including scalability to much larger training sets, more efficient use of the per-expression token budget, and systematic training set decontamination with respect to equivalent test expressions. We demonstrate these advantages in our Flash-ANSR framework, which achieves much better accuracy than amortized baselines (NeSymReS, E2E) on the FastSRB benchmark. Moreover, it performs on par with state-of-the-art direct optimization (PySR) while recovering more concise instead of more complex expressions with increasing inference budget.
Authors: Fatemeh Nejati, Mahdi Rabbani, Morteza Eskandarian, Mansur Mirani, Gunjan Piya, Igor Opushnyev, Ali A. Ghorbani, Sajjad Dadkhah
Abstract: Phishing attacks represents one of the primary attack methods which is used by cyber attackers. In many cases, attackers use deceptive emails along with malicious attachments to trick users into giving away sensitive information or installing malware while compromising entire systems. The flexibility of malicious email attachments makes them stand out as a preferred vector for attackers as they can embed harmful content such as malware or malicious URLs inside standard document formats. Although phishing email defenses have improved a lot, attackers continue to abuse attachments, enabling malicious content to bypass security measures. Moreover, another challenge that researches face in training advance models, is lack of an unified and comprehensive dataset that covers the most prevalent data types. To address this gap, we generated CIC-Trap4Phish, a multi-format dataset containing both malicious and benign samples across five categories commonly used in phishing campaigns: Microsoft Word documents, Excel spreadsheets, PDF files, HTML pages, and QR code images. For the first four file types, a set of execution-free static feature pipeline was proposed, designed to capture structural, lexical, and metadata-based indicators without the need to open or execute files. Feature selection was performed using a combination of SHAP analysis and feature importance, yielding compact, discriminative feature subsets for each file type. The selected features were evaluated by using lightweight machine learning models, including Random Forest, XGBoost, and Decision Tree. All models demonstrate high detection accuracy across formats. For QR code-based phishing (quishing), two complementary methods were implemented: image-based detection by employing Convolutional Neural Networks (CNNs) and lexical analysis of decoded URLs using recent lightweight language models.
Authors: Djohan Bonnet, Jamie Lohoff, Jan Finkbeiner, Elidona Skhikerujah, Emre Neftci
Abstract: In-Context Learning (ICL) in transformers acts as an online associative memory and is believed to underpin their high performance on complex sequence processing tasks. However, in gated linear attention models, this memory has a fixed capacity and is prone to interference, especially for long sequences. We propose Palimpsa, a self-attention model that views ICL as a continual learning problem that must address a stability-plasticity dilemma. Palimpsa uses Bayesian metaplasticity, where the plasticity of each attention state is tied to an importance state grounded by a prior distribution that captures accumulated knowledge. We demonstrate that various gated linear attention models emerge as specific architecture choices and posterior approximations, and that Mamba2 is a special case of Palimpsa where forgetting dominates. This theoretical link enables the transformation of any non-metaplastic model into a metaplastic one, significantly expanding its memory capacity. Our experiments show that Palimpsa consistently outperforms baselines on the Multi-Query Associative Recall (MQAR) benchmark and on Commonsense Reasoning tasks.
Authors: Zhirui Zhang, Hongbo Zhang, Haoxiang Fei, Zhiyuan Bao, Yubin Chen, Zhengyu Lei, Ziyue Liu, Yixuan Sun, Mingkun Xiao, Zihang Ye, Yu Zhang, Hongcheng Zhu, Yuxiang Wen, Heung-Yeung Shum
Abstract: Although large language models (LLMs) have demonstrated impressive coding capabilities, their ability to autonomously build production-scale software from explicit specifications remains an open question. We introduce SWE-AGI, an open-source benchmark for evaluating end-to-end, specification-driven construction of software systems written in MoonBit. SWE-AGI tasks require LLM-based agents to implement parsers, interpreters, binary decoders, and SAT solvers strictly from authoritative standards and RFCs under a fixed API scaffold. Each task involves implementing 1,000-10,000 lines of core logic, corresponding to weeks or months of engineering effort for an experienced human developer. By leveraging the nascent MoonBit ecosystem, SWE-AGI minimizes data leakage, forcing agents to rely on long-horizon architectural reasoning rather than code retrieval. Across frontier models, gpt-5.3-codex achieves the best overall performance (solving 19/22 tasks, 86.4%), outperforming claude-opus-4.6 (15/22, 68.2%), and kimi-2.5 exhibits the strongest performance among open-source models. Performance degrades sharply with increasing task difficulty, particularly on hard, specification-intensive systems. Behavioral analysis further reveals that as codebases scale, code reading, rather than writing, becomes the dominant bottleneck in AI-assisted development. Overall, while specification-driven autonomous software engineering is increasingly viable, substantial challenges remain before it can reliably support production-scale development.
Authors: Xavier Hu, Jinxiang Xia, Shengze Xu, Kangqi Song, Yishuo Yuan, Guibin Zhang, JinCheng Ren, Boyu Feng, Li Lu, Tieyong Zeng, Jiaheng Liu, Minghao Liu, He Zhu, Yuchen Eleanor Jiang, Wei Wang, Wangchunshu Zhou
Abstract: Long-horizon planning is widely recognized as a core capability of autonomous LLM-based agents; however, current evaluation frameworks suffer from being largely episodic, domain-specific, or insufficiently grounded in persistent economic dynamics. We introduce EcoGym, a generalizable benchmark for continuous plan-and-execute decision making in interactive economies. EcoGym comprises three diverse environments: Vending, Freelance, and Operation, implemented in a unified decision-making process with standardized interfaces, and budgeted actions over an effectively unbounded horizon (1000+ steps if 365 day-loops for evaluation). The evaluation of EcoGym is based on business-relevant outcomes (e.g., net worth, income, and DAU), targeting long-term strategic coherence and robustness under partial observability and stochasticity. Experiments across eleven leading LLMs expose a systematic tension: no single model dominates across all three scenarios. Critically, we find that models exhibit significant suboptimality in either high-level strategies or efficient actions executions. EcoGym is released as an open, extensible testbed for transparent long-horizon agent evaluation and for studying controllability-utility trade-offs in realistic economic settings.
Authors: Daisuke Nohara, Taishi Nakamura, Rio Yokota
Abstract: Reinforcement learning substantially improves reasoning in large language models, but it also tends to lengthen chain of thought outputs and increase computational cost during both training and inference. Though length control methods have been proposed, it remains unclear what the optimal output length is for balancing efficiency and performance. In this work, we compare several length control methods on two models, Qwen3-1.7B Base and DeepSeek-R1-Distill-Qwen-1.5B. Our results indicate that length penalties may hinder reasoning acquisition, while properly tuned length control can improve efficiency for models with strong prior reasoning. By extending prior work to RL trained policies, we identify two failure modes, 1) long outputs increase dispersion, and 2) short outputs lead to under-thinking.
Authors: Azkaa Nasir, Fatima Dossa, Muhammad Ahmed Atif, Mohammad Shahid Shaikh
Abstract: Transfer learning in deep reinforcement learning is often motivated by improved stability and reduced training cost, but it can also fail under substantial domain shift. This paper presents a controlled empirical study examining how architectural differences between Double Deep Q-Networks (DDQN) and Dueling DQN influence transfer behavior across environments. Using CartPole as a source task and LunarLander as a structurally distinct target task, we evaluate a fixed layer-wise representation transfer protocol under identical hyperparameters and training conditions, with baseline agents trained from scratch used to contextualize transfer effects. Empirical results show that DDQN consistently avoids negative transfer under the examined setup and maintains learning dynamics comparable to baseline performance in the target environment. In contrast, Dueling DQN consistently exhibits negative transfer under identical conditions, characterized by degraded rewards and unstable optimization behavior. Statistical analysis across multiple random seeds confirms a significant performance gap under transfer. These findings suggest that architectural inductive bias is strongly associated with robustness to cross-environment transfer in value-based deep reinforcement learning under the examined transfer protocol.
Authors: Jiaquan Zhang, Chaoning Zhang, Shuxu Chen, Yibei Liu, Chenghao Li, Qigan Sun, Shuai Yuan, Fachrina Dewi Puspitasari, Dongshen Han, Guoqing Wang, Sung-Ho Bae, Yang Yang
Abstract: Text summarization is a fundamental task in natural language processing (NLP), and the information explosion has made long-document processing increasingly demanding, making summarization essential. Existing research mainly focuses on model improvements and sentence-level pruning, but often overlooks global structure, leading to disrupted coherence and weakened downstream performance. Some studies employ large language models (LLMs), which achieve higher accuracy but incur substantial resource and time costs. To address these issues, we introduce GloSA-sum, the first summarization approach that achieves global structure awareness via topological data analysis (TDA). GloSA-sum summarizes text efficiently while preserving semantic cores and logical dependencies. Specifically, we construct a semantic-weighted graph from sentence embeddings, where persistent homology identifies core semantics and logical structures, preserved in a ``protection pool'' as the backbone for summarization. We design a topology-guided iterative strategy, where lightweight proxy metrics approximate sentence importance to avoid repeated high-cost computations, thus preserving structural integrity while improving efficiency. To further enhance long-text processing, we propose a hierarchical strategy that integrates segment-level and global summarization. Experiments on multiple datasets demonstrate that GloSA-sum reduces redundancy while preserving semantic and logical integrity, striking a balance between accuracy and efficiency, and further benefits LLM downstream tasks by shortening contexts while retaining essential reasoning chains.
Authors: Zongrui Li, Xinhua Ma, Minghui Hu, Yunqing Zhao, Yingchen Yu, Qian Zheng, Chang Liu, Xudong Jiang, Song Bai
Abstract: Monocular normal estimation aims to estimate the normal map from a single RGB image of an object under arbitrary lights. Existing methods rely on deep models to directly predict normal maps. However, they often suffer from 3D misalignment: while the estimated normal maps may appear to have a correct appearance, the reconstructed surfaces often fail to align with the geometric details. We argue that this misalignment stems from the current paradigm: the model struggles to distinguish and reconstruct varying geometry represented in normal maps, as the differences in underlying geometry are reflected only through relatively subtle color variations. To address this issue, we propose a new paradigm that reformulates normal estimation as shading sequence estimation, where shading sequences are more sensitive to various geometric information. Building on this paradigm, we present RoSE, a method that leverages image-to-video generative models to predict shading sequences. The predicted shading sequences are then converted into normal maps by solving a simple ordinary least-squares problem. To enhance robustness and better handle complex objects, RoSE is trained on a synthetic dataset, MultiShade, with diverse shapes, materials, and light conditions. Experiments demonstrate that RoSE achieves state-of-the-art performance on real-world benchmark datasets for object-based monocular normal estimation.
Authors: J Rosser, Robert Kirk, Edward Grefenstette, Jakob Foerster, Laura Ruis
Abstract: Influence functions are commonly used to attribute model behavior to training documents. We explore the reverse: crafting training data that induces model behavior. Our framework, Infusion, uses scalable influence-function approximations to compute small perturbations to training documents that induce targeted changes in model behavior through parameter shifts. We evaluate Infusion on data poisoning tasks across vision and language domains. On CIFAR-10, we show that making subtle edits via Infusion to just 0.2% (100/45,000) of the training documents can be competitive with the baseline of inserting a small number of explicit behavior examples. We also find that Infusion transfers across architectures (ResNet $\leftrightarrow$ CNN), suggesting a single poisoned corpus can affect multiple independently trained models. In preliminary language experiments, we characterize when our approach increases the probability of target behaviors and when it fails, finding it most effective at amplifying behaviors the model has already learned. Taken together, these results show that small, subtle edits to training data can systematically shape model behavior, underscoring the importance of training data interpretability for adversaries and defenders alike. We provide the code here: https://github.com/jrosseruk/infusion.
Authors: Dharmendra Sharma, Archit Sharma, John Rebeiro, Vaibhav Kesharwani, Peeyush Thakur, Narendra Kumar Dhar, Laxmidhar Behera
Abstract: Temporally locating and classifying fine-grained sub-task segments in long, untrimmed videos is crucial to safe human-robot collaboration. Unlike generic activity recognition, collaborative manipulation requires sub-task labels that are directly robot-executable. We present RoboSubtaskNet, a multi-stage human-to-robot sub-task segmentation framework that couples attention-enhanced I3D features (RGB plus optical flow) with a modified MS-TCN employing a Fibonacci dilation schedule to capture better short-horizon transitions such as reach-pick-place. The network is trained with a composite objective comprising cross-entropy and temporal regularizers (truncated MSE and a transition-aware term) to reduce over-segmentation and to encourage valid sub-task progressions. To close the gap between vision benchmarks and control, we introduce RoboSubtask, a dataset of healthcare and industrial demonstrations annotated at the sub-task level and designed for deterministic mapping to manipulator primitives. Empirically, RoboSubtaskNet outperforms MS-TCN and MS-TCN++ on GTEA and our RoboSubtask benchmark (boundary-sensitive and sequence metrics), while remaining competitive on the long-horizon Breakfast benchmark. Specifically, RoboSubtaskNet attains F1 @ 50 = 79.5%, Edit = 88.6%, Acc = 78.9% on GTEA; F1 @ 50 = 30.4%, Edit = 52.0%, Acc = 53.5% on Breakfast; and F1 @ 50 = 94.2%, Edit = 95.6%, Acc = 92.2% on RoboSubtask. We further validate the full perception-to-execution pipeline on a 7-DoF Kinova Gen3 manipulator, achieving reliable end-to-end behavior in physical trials (overall task success approx 91.25%). These results demonstrate a practical path from sub-task level video understanding to deployed robotic manipulation in real-world settings.
Authors: Changjiang Jiang, Xinkuan Sha, Fengchang Yu, Jingjing Liu, Jian Liu, Mingqi Fang, Chenfeng Zhang, Wei Lu
Abstract: Recent studies have demonstrated that incorporating Chain-of-Thought (CoT) reasoning into the detection process can enhance a model's ability to detect synthetic images. However, excessively lengthy reasoning incurs substantial resource overhead, including token consumption and latency, which is particularly redundant when handling obviously generated forgeries. To address this issue, we propose Fake-HR1, a large-scale hybrid-reasoning model that, to the best of our knowledge, is the first to adaptively determine whether reasoning is necessary based on the characteristics of the generative detection task. To achieve this, we design a two-stage training framework: we first perform Hybrid Fine-Tuning (HFT) for cold-start initialization, followed by online reinforcement learning with Hybrid-Reasoning Grouped Policy Optimization (HGRPO) to implicitly learn when to select an appropriate reasoning mode. Experimental results show that Fake-HR1 adaptively performs reasoning across different types of queries, surpassing existing LLMs in both reasoning ability and generative detection performance, while significantly improving response efficiency.