arXiv:2601.06453v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly grounded in sensor data to perceive and reason about human physiology and the physical world. However, accurately interpreting heterogeneous multimodal sensor data remains a fundamental challenge. We show that a single monolithic LLM often fails to reason coherently across modalities, leading to incomplete interpretations and prior-knowledge bias. We introduce ConSensus, a training-free multi-agent collaboration framework that decomposes multimodal sensing tasks into specialized, modality-aware agents. To aggregate agent-level interpretations, we propose a hybrid fusion mechanism that balances semantic aggregation, which enables cross-modal reasoning and contextual understanding, with statistical consensus, which provides robustness through agreement across modalities. While each approach has complementary failure modes, their combination enables reliable inference under sensor noise and missing data. We evaluate ConSensus on five diverse multimodal sensing benchmarks, demonstrating an average accuracy improvement of 7.1% over the single-agent baseline. Furthermore, ConSensus matches or exceeds the performance of iterative multi-agent debate methods while achieving a 12.7 times reduction in average fusion token cost through a single-round hybrid fusion protocol, yielding a robust and efficient solution for real-world multimodal sensing tasks.

arXiv:2601.06663v1 Announce Type: new Abstract: Large language model-based agents are rapidly evolving from simple conversational assistants into autonomous systems capable of performing complex, professional-level tasks in various domains. While these advancements promise significant productivity gains, they also introduce critical safety risks that remain under-explored. Existing safety evaluations primarily focus on simple, daily assistance tasks, failing to capture the intricate decision-making processes and potential consequences of misaligned behaviors in professional settings. To address this gap, we introduce \textbf{SafePro}, a comprehensive benchmark designed to evaluate the safety alignment of AI agents performing professional activities. SafePro features a dataset of high-complexity tasks across diverse professional domains with safety risks, developed through a rigorous iterative creation and review process. Our evaluation of state-of-the-art AI models reveals significant safety vulnerabilities and uncovers new unsafe behaviors in professional contexts. We further show that these models exhibit both insufficient safety judgment and weak safety alignment when executing complex professional tasks. In addition, we investigate safety mitigation strategies for improving agent safety in these scenarios and observe encouraging improvements. Together, our findings highlight the urgent need for robust safety mechanisms tailored to the next generation of professional AI agents.

arXiv:2601.07463v1 Announce Type: new Abstract: Offline multi-agent reinforcement learning (MARL) aims to solve cooperative decision-making problems in multi-agent systems using pre-collected datasets. Existing offline MARL methods primarily constrain training within the dataset distribution, resulting in overly conservative policies that struggle to generalize beyond the support of the data. While model-based approaches offer a promising solution by expanding the original dataset with synthetic data generated from a learned world model, the high dimensionality, non-stationarity, and complexity of multi-agent systems make it challenging to accurately estimate the transitions and reward functions in offline MARL. Given the difficulty of directly modeling joint dynamics, we propose a local-to-global (LOGO) world model, a novel framework that leverages local predictions-which are easier to estimate-to infer global state dynamics, thus improving prediction accuracy while implicitly capturing agent-wise dependencies. Using the trained world model, we generate synthetic data to augment the original dataset, expanding the effective state-action space. To ensure reliable policy learning, we further introduce an uncertainty-aware sampling mechanism that adaptively weights synthetic data by prediction uncertainty, reducing approximation error propagation to policies. In contrast to conventional ensemble-based methods, our approach requires only an additional encoder for uncertainty estimation, significantly reducing computational overhead while maintaining accuracy. Extensive experiments across 8 scenarios against 8 baselines demonstrate that our method surpasses state-of-the-art baselines on standard offline MARL benchmarks, establishing a new model-based baseline for generalizable offline multi-agent learning.

arXiv:2601.06060v1 Announce Type: cross Abstract: AI-generated "slop" is often seen as digital pollution. We argue that this dismissal of the topic risks missing important aspects of AI Slop that deserve rigorous study. AI Slop serves a social function: it offers a supply-side solution to a variety of problems in cultural and economic demand - that, collectively, people want more content than humans can supply. We also argue that AI Slop is not mere digital detritus but has its own aesthetic value. Like other "low" cultural forms initially dismissed by critics, it nonetheless offers a legitimate means of collective sense-making, with the potential to express meaning and identity. We identify three key features of family resemblance for prototypical AI Slop: superficial competence (its veneer of quality is belied by a deeper lack of substance), asymmetry effort (it takes vastly less effort to generate than would be the case without AI), and mass producibility (it is part of a digital ecosystem of widespread generation and consumption). While AI Slop is heterogeneous and depends crucially on its medium, it tends to vary across three dimensions: instrumental utility, personalization, and surrealism. AI Slop will be an increasingly prolific and impactful part of our creative, information, and cultural economies; we should take it seriously as an object of study in its own right.

arXiv:2601.07344v1 Announce Type: cross Abstract: Recent advances in medical multi-modal models focus on specialized image analysis like dermatology, pathology, or radiology. However, they do not fully capture the complexity of real-world clinical diagnostics, which involve heterogeneous inputs and require ongoing contextual understanding during patient-physician interactions. To bridge this gap, we introduce PulseMind, a new family of multi-modal diagnostic models that integrates a systematically curated dataset, a comprehensive evaluation benchmark, and a tailored training framework. Specifically, we first construct a diagnostic dataset, MediScope, which comprises 98,000 real-world multi-turn consultations and 601,500 medical images, spanning over 10 major clinical departments and more than 200 sub-specialties. Then, to better reflect the requirements of real-world clinical diagnosis, we develop the PulseMind Benchmark, a multi-turn diagnostic consultation benchmark with a four-dimensional evaluation protocol comprising proactiveness, accuracy, usefulness, and language quality. Finally, we design a training framework tailored for multi-modal clinical diagnostics, centered around a core component named Comparison-based Reinforcement Policy Optimization (CRPO). Compared to absolute score rewards, CRPO uses relative preference signals from multi-dimensional com-parisons to provide stable and human-aligned training guidance. Extensive experiments demonstrate that PulseMind achieves competitive performance on both the diagnostic consultation benchmark and public medical benchmarks.

arXiv:2505.19562v2 Announce Type: replace Abstract: Large language models (LLMs) are approaching expert-level performance in medical question answering (QA), demonstrating strong potential to improve public healthcare. However, underlying biases related to sensitive attributes such as sex and race pose life-critical risks. The extent to which such sensitive attributes affect diagnosis remains an open question and requires comprehensive empirical investigation. Additionally, even the latest Counterfactual Patient Variations (CPV) benchmark can hardly distinguish the bias levels of different LLMs. To further explore these dynamics, we propose a new benchmark, FairMedQA, and benchmark 12 representative LLMs. FairMedQA contains 4,806 counterfactual question pairs constructed from 801 clinical vignettes. Our results reveal substantial accuracy disparity ranging from 3 to 19 percentage points across sensitive demographic groups. Notably, FairMedQA exposes biases that are at least 12 percentage points larger than those identified by the latest CPV benchmark, presenting superior benchmarking sensitivity. Our results underscore an urgent need for targeted debiasing techniques and more rigorous, identity-aware validation protocols before LLMs can be safely integrated into practical clinical decision-support systems.

arXiv:2512.20387v3 Announce Type: replace Abstract: We propose a Vision-Language Simulation Model (VLSM) that unifies visual and textual understanding to synthesize executable FlexScript from layout sketches and natural-language prompts, enabling cross-modal reasoning for industrial simulation systems. To support this new paradigm, the study constructs the first large-scale dataset for generative digital twins, comprising over 120,000 prompt-sketch-code triplets that enable multimodal learning between textual descriptions, spatial structures, and simulation logic. In parallel, three novel evaluation metrics, Structural Validity Rate (SVR), Parameter Match Rate (PMR), and Execution Success Rate (ESR), are proposed specifically for this task to comprehensively evaluate structural integrity, parameter fidelity, and simulator executability. Through systematic ablation across vision encoders, connectors, and code-pretrained language backbones, the proposed models achieve near-perfect structural accuracy and high execution robustness. This work establishes a foundation for generative digital twins that integrate visual reasoning and language understanding into executable industrial simulation systems. Project page: https://danielhsu2014.github.io/GDT-VLSM-project/

Nature Medicine, Published online: 12 January 2026; doi:10.1038/s41591-025-04126-3

Including data from 1,047 patients across 19 inflammatory diseases, a new atlas presents a comprehensive model of inflammation in circulating immune cells.

arXiv:2601.05529v1 Announce Type: new Abstract: One mistake by an AI system in a safety-critical setting can cost lives. As Large Language Models (LLMs) become integral to robotics decision-making, the physical dimension of risk grows; a single wrong instruction can directly endanger human safety. This paper addresses the urgent need to systematically evaluate LLM performance in scenarios where even minor errors are catastrophic. Through a qualitative evaluation of a fire evacuation scenario, we identified critical failure cases in LLM-based decision-making. Based on these, we designed seven tasks for quantitative assessment, categorized into: Complete Information, Incomplete Information, and Safety-Oriented Spatial Reasoning (SOSR). Complete information tasks utilize ASCII maps to minimize interpretation ambiguity and isolate spatial reasoning from visual processing. Incomplete information tasks require models to infer missing context, testing for spatial continuity versus hallucinations. SOSR tasks use natural language to evaluate safe decision-making in life-threatening contexts. We benchmark various LLMs and Vision-Language Models (VLMs) across these tasks. Beyond aggregate performance, we analyze the implications of a 1% failure rate, highlighting how "rare" errors escalate into catastrophic outcomes. Results reveal serious vulnerabilities: several models achieved a 0% success rate in ASCII navigation, while in a simulated fire drill, models instructed robots to move toward hazardous areas instead of emergency exits. Our findings lead to a sobering conclusion: current LLMs are not ready for direct deployment in safety-critical systems. A 99% accuracy rate is dangerously misleading in robotics, as it implies one out of every hundred executions could result in catastrophic harm. We demonstrate that even state-of-the-art models cannot guarantee safety, and absolute reliance on them creates unacceptable risks.

Nature Medicine, Published online: 09 January 2026; doi:10.1038/s41591-025-04170-z

Analysis of a placebo-controlled trial of a BCMA-targeting CAR-T cell therapy in patients with myasthenia gravis shows that CAR-T cell infusion selectively remodels the systemic immune environment, with elimination of BCMA-high plasma cells and activated plasmacytoid dendritic cells and changes in the autoreactive B cell repertoire.

arXiv:2601.04506v1 Announce Type: cross Abstract: Therapeutic peptides show promise in targeting previously undruggable binding sites, with recent advancements in deep generative models enabling full-atom peptide co-design for specific protein receptors. However, the critical role of molecular surfaces in protein-protein interactions (PPIs) has been underexplored. To bridge this gap, we propose an omni-design peptides generation paradigm, called SurfFlow, a novel surface-based generative algorithm that enables comprehensive co-design of sequence, structure, and surface for peptides. SurfFlow employs a multi-modality conditional flow matching (CFM) architecture to learn distributions of surface geometries and biochemical properties, enhancing peptide binding accuracy. Evaluated on the comprehensive PepMerge benchmark, SurfFlow consistently outperforms full-atom baselines across all metrics. These results highlight the advantages of considering molecular surfaces in de novo peptide discovery and demonstrate the potential of integrating multiple protein modalities for more effective therapeutic peptide discovery.

arXiv:2601.04790v1 Announce Type: cross Abstract: Multi-agent systems utilizing large language models often assign authoritative roles to improve performance, yet the impact of authority bias on agent interactions remains underexplored. We present the first systematic analysis of role-based authority bias in free-form multi-agent evaluation using ChatEval. Applying French and Raven's power-based theory, we classify authoritative roles into legitimate, referent, and expert types and analyze their influence across 12-turn conversations. Experiments with GPT-4o and DeepSeek R1 reveal that Expert and Referent power roles exert stronger influence than Legitimate power roles. Crucially, authority bias emerges not through active conformity by general agents, but through authoritative roles consistently maintaining their positions while general agents demonstrate flexibility. Furthermore, authority influence requires clear position statements, as neutral responses fail to generate bias. These findings provide key insights for designing multi-agent frameworks with asymmetric interaction patterns.

arXiv:2601.05148v1 Announce Type: cross Abstract: Pathology foundation models substantially advanced the possibilities in computational pathology -- yet tradeoffs in terms of performance, robustness, and computational requirements remained, which limited their clinical deployment. In this report, we present Atlas 2, Atlas 2-B, and Atlas 2-S, three pathology vision foundation models which bridge these shortcomings by showing state-of-the-art performance in prediction performance, robustness, and resource efficiency in a comprehensive evaluation across eighty public benchmarks. Our models were trained on the largest pathology foundation model dataset to date comprising 5.5 million histopathology whole slide images, collected from three medical institutions Charit\'e - Universt\"atsmedizin Berlin, LMU Munich, and Mayo Clinic.

arXiv:2601.01802v3 Announce Type: replace Abstract: To develop a reliable AI for psychological assessment, we introduce \texttt{PsychEval}, a multi-session, multi-therapy, and highly realistic benchmark designed to address three key challenges: \textbf{1) Can we train a highly realistic AI counselor?} Realistic counseling is a longitudinal task requiring sustained memory and dynamic goal tracking. We propose a multi-session benchmark (spanning 6-10 sessions across three distinct stages) that demands critical capabilities such as memory continuity, adaptive reasoning, and longitudinal planning. The dataset is annotated with extensive professional skills, comprising over 677 meta-skills and 4577 atomic skills. \textbf{2) How to train a multi-therapy AI counselor?} While existing models often focus on a single therapy, complex cases frequently require flexible strategies among various therapies. We construct a diverse dataset covering five therapeutic modalities (Psychodynamic, Behaviorism, CBT, Humanistic Existentialist, and Postmodernist) alongside an integrative therapy with a unified three-stage clinical framework across six core psychological topics. \textbf{3) How to systematically evaluate an AI counselor?} We establish a holistic evaluation framework with 18 therapy-specific and therapy-shared metrics across Client-Level and Counselor-Level dimensions. To support this, we also construct over 2,000 diverse client profiles. Extensive experimental analysis fully validates the superior quality and clinical fidelity of our dataset. Crucially, \texttt{PsychEval} transcends static benchmarking to serve as a high-fidelity reinforcement learning environment that enables the self-evolutionary training of clinically responsible and adaptive AI counselors.

Background: The rapid growth of user-generated web-based health information increases the complexity of cancer information seeking. One promising strategy for promoting high-quality cancer information consumption is through targeted interventions that are intentionally designed to reach individuals in the web-based spaces they occupy. However, there is a paucity of evidence-based information on the best strategies for designing and implementing web-based health behavior change interventions to improve individuals’ cancer-related knowledge and prevent cancer. Objective: This study aimed to develop and pilot test a theory-based intervention via the web to reduce 6 cancer risk factors among rural emerging adults (EAs) through community-engaged research. Methods: This mixed methods evaluation describes the development of a web-based cancer prevention intervention aimed at rural EAs aged 18-26 years in the United States and delivered in Facebook private groups. The intervention was guided by behavior change theory and cocreated with EA and Stakeholder Organization Advisory Boards to ensure relevance, accessibility, and appropriateness. We report on 3 formative surveys, a pilot intervention, protocol development, and the community-engaged process for intervention development. Descriptive statistics were applied to the surveys and pilot intervention baseline results to produce means and SDs using R. Results: We developed posts (n=400) for a Facebook feed aimed at reducing 6 cancer risk behaviors (unhealthy diet, lack of physical activity, tobacco use, alcohol use, sun exposure, and human papillomavirus infection) with iterative input from the EA and stakeholder advisory boards. Formative surveys with rural EAs (n=297) and a pilot study of the intervention with this population (n=26) were conducted. In the pilot study, the intervention reached participants across rural counties, with sustained engagement (post views=1060, reactions=346, comments=72) over a one-month period. Key modifications to the intervention content and design emerged from both advisory boards, the formative surveys, and the pilot intervention, focusing on using perceived reliable sources and direct links to source material. Conclusions: This web-based cancer prevention intervention is scalable and delivers engaging, evidence-informed health information to rural EAs. We offer key insights into the design and implementation of web-based cancer prevention interventions for EAs by describing the resources, timelines, and expertise needed to design and implement the intervention. Considerations for fully engaging EA and community stakeholder partners are presented, and we discuss how their involvement resulted in modifications that strengthened the intervention. Finally, we highlight the importance of theory-based health-behavior messaging, digital messaging skillsets, and platform-tailored dissemination strategies for maximizing web-based intervention acceptability. Trial Registration: ClinicalTrials.gov NCT05618158; https://classic.clinicaltrials.gov/ct2/show/NCT05618158

npj Digital Medicine, Published online: 07 January 2026; doi:10.1038/s41746-025-02324-4

An autonomous agentic workflow for clinical detection of cognitive concerns using large language models

Background: Gastroesophageal reflux disease (GERD) symptoms significantly affect patients’ quality of life. Patient-reported outcome (PRO) instruments for symptoms measurement in GERD patients is advocated by regulatory authority. Current tools for GERD symptoms evaluation are limited and the results can be biased by the recall bias. To better characterize the GERD symptoms, an e-diary was developed for daily GERD symptom monitoring. Objective: To build up and optimize a PRO-based e-diary, and to investigate the effect of symptom frequency on adherence. Methods: The GERD e-diary evaluated 8 daytime (acid regurgitation, cough, heartburn, sour taste in the mouth, hiccups, hoarseness, dysphagia, and chest pain) and 2 nighttime symptoms (acid regurgitation and cough) for consecutive 8 weeks. The adherence of e-diary, defined as daily completing rate of e-diary, was evaluated and optimized from First Stage to Third Stage with no reminder implemented in First Stage, sending reminding SMS (Short Message Service) text messaging upon detecting missing data in Second Stage (no reminder during the first 3 to 5 days after enrollment), and immediate installation of reminding system at enrollment in Third Stage. GERD symptom frequency was obtained by summation of the symptomatic days in each week. A multiple regression analysis was performed to examine the effects of system optimization and GERD symptom frequency on patient adherence, while controlling for potential confounding variables. Results: 138 GERD patients (M/F=70/68; age: mean 52.9, SD 12.3 years) were recruited. At First Stage, the adherence was 47.2%, 40% and 57.6% for nighttime, daytime and overall symptom. System optimization significantly improved adherence with increased adherence of nighttime symptoms by 12.5% (P=.005) and 10.9% (P=.01), daytime symptom by 21.7% (P

arXiv:2601.01576v1 Announce Type: cross Abstract: Evaluating novelty is critical yet challenging in peer review, as reviewers must assess submissions against a vast, rapidly evolving literature. This report presents OpenNovelty, an LLM-powered agentic system for transparent, evidence-based novelty analysis. The system operates through four phases: (1) extracting the core task and contribution claims to generate retrieval queries; (2) retrieving relevant prior work based on extracted queries via semantic search engine; (3) constructing a hierarchical taxonomy of core-task-related work and performing contribution-level full-text comparisons against each contribution; and (4) synthesizing all analyses into a structured novelty report with explicit citations and evidence snippets. Unlike naive LLM-based approaches, \textsc{OpenNovelty} grounds all assessments in retrieved real papers, ensuring verifiable judgments. We deploy our system on 500+ ICLR 2026 submissions with all reports publicly available on our website, and preliminary analysis suggests it can identify relevant prior work, including closely related papers that authors may overlook. OpenNovelty aims to empower the research community with a scalable tool that promotes fair, consistent, and evidence-backed peer review.

arXiv:2601.01627v1 Announce Type: cross Abstract: As Large Language Models (LLMs) are increasingly deployed in healthcare field, it becomes essential to carefully evaluate their medical safety before clinical use. However, existing safety benchmarks remain predominantly English-centric, and test with only single-turn prompts despite multi-turn clinical consultations. To address these gaps, we introduce JMedEthicBench, the first multi-turn conversational benchmark for evaluating medical safety of LLMs for Japanese healthcare. Our benchmark is based on 67 guidelines from the Japan Medical Association and contains over 50,000 adversarial conversations generated using seven automatically discovered jailbreak strategies. Using a dual-LLM scoring protocol, we evaluate 27 models and find that commercial models maintain robust safety while medical-specialized models exhibit increased vulnerability. Furthermore, safety scores decline significantly across conversation turns (median: 9.5 to 5.0, $p