❌

Reading view

↗️

JarvisGUI: Towards Cross-Device GUI Agents with Dynamic Task Composition

✇cs.AI, q-bio.NC updates on arXiv.org
By: Zixiang Chen · Yuheng Lu · Zihao Cheng · Zeming Liu · Jizeng Bai · Ziye Huang · Zhiyin Lin · Zihan Li · Yuhang Guo · Yunhong Wang · Haifeng Wang
arXiv:2609.10451v1 Announce Type: new Abstract: Real-world GUI usage frequently involves workflows that span multiple devices and platforms, requiring the transfer of intermediate results, maintenance of shared state, and coordination across heterogeneous environments. However, existing GUI benchmarks overwhelmingly evaluate agents on single-device, statically defined tasks, thus leaving such cross-device capabilities largely unexamined, resulting in an overly optimistic assessment of agents' readiness for real-world usage. We introduce JarvisGUI, a dynamic benchmark that evaluates GUI agents on cross-device workflows requiring coordinated interaction across heterogeneous platforms, including Android, Windows, and Ubuntu. Specifically, JarvisGUI formulates GUI tasks as input-output transformations under a lightweight type system, which allows us to automatically compose multi-step, cross-device workflows and dynamically evaluate agent performance within a unified framework. By evaluating agents in virtual environments spanning multiple operating systems, JarvisGUI reveals that state-of-the-art open-source GUI agents struggle with the state-transfer awareness, cross-platform contextual reasoning, and long-horizon dependency management required for real-world workflows, exposing a critical capability gap invisible to existing benchmarks.
  •  
  • ↗️
↗️

Advances in Radiomics for Immune Checkpoint Inhibitor-related Pneumonitis of Lung Cancer

✇Omics In Lung
By: Dong Zhang · Li Peng · Tianming Zhang · Fanqi Wu · Hong Wang

Zhongguo Fei Ai Za Zhi. 2026 Jul 20;29(7):540-547. doi: 10.3779/j.issn.1009-3419.2026.101.17.

ABSTRACT

Immune checkpoint inhibitors (ICIs) have significantly improved the prognosis of patients with lung cancer. However, checkpoint inhibitor-related pneumonitis (CIP), as one of the most severe immune-related adverse events, lacks well-defined diagnostic criteria and reliable risk stratification tools. Radiomics enables high-throughput feature extraction from computed tomography images and provides a non-invasive technical approach for the early identification and risk stratification of CIP. This article systematically reviews the recent advances in the application of radiomics to risk prediction, diagnosis and differential diagnosis, and prognostic evaluation of CIP in lung cancer immunotherapy. Furthermore, it explores the value of integrating radiomics with multi-omics data in elucidating the pathogenesis of CIP, as well as the role of explainable artificial intelligence (XAI) in enhancing the clinical trustworthiness of models. .

PMID:42705857 | DOI:10.3779/j.issn.1009-3419.2026.101.17

  •  
  • ↗️
↗️

Circadian-based individualised protection against inflammation-cancer transition in atrophic gastritis patients

✇Omics in Gastric
By: Ceng Shan · Min Guo · Feifan Feng · Xihong Wang · Mohang Li · Weichao Xu · Keyu Jia · Yifei Liu · Yifan Jia

EPMA J. 2026 Aug 21;17(3):665-700. doi: 10.1007/s13167-026-00465-4. eCollection 2026 Sep.

ABSTRACT

Chronic atrophic gastritis (CAG) is a critical precancerous stage in the development of gastric cancer (GC). Circadian rhythm disruption perturbs the core clock gene network, including circadian locomotor output cycles kaput (CLOCK), brain and muscle ARNT-like 1 (BMAL1), period circadian protein homolog (PER), and cryptochrome (CRY). These alterations contribute to a multi-layered pathological cascade involving DNA damage accumulation, epigenetic remodeling, altered epithelial cell plasticity, cellular senescence, microbiota dysbiosis, tumor microenvironment remodeling, metabolic reprogramming, aberrant angiogenesis, and dysregulated cell death, thereby accelerating CAG to GC progression. However, existing studies have predominantly treated the circadian rhythm as a passive risk factor for disease onset and have yet to elevate it to an actionable interventional target within the full-course management of gastric precancerous lesions. Building on a systematic synthesis of the mechanistic evidence outlined above, this review proposes a predictive, preventive and personalised medicine (PPPM/3PM) three-tier management framework grounded in circadian-based individualised protection. At the predictive level, digital biomarkers (sleep-wake rhythms, light exposure, physical activity, and dietary behavior), multi-omics profiles, and circadian-related molecular signatures are integrated to achieve dynamic risk stratification of CAG populations. At the targeted prevention level, pharmacological agents and natural compounds with circadian-regulating potential are deployed to develop proactive protective strategies tailored to distinct pathological stages and circadian phenotypes. At the personalised treatment level, lifestyle interventions, chronotherapy, nano-carrier-based circadian-synchronised delivery, and dynamic biomarker monitoring are combined to formulate precision intervention regimens informed by individual circadian phenotypes. This framework repositions the circadian rhythm from a latent risk factor to a protectable and therapeutically targetable axis, offering new insights into time-optimised intervention strategies for the inflammation to cancer transition in CAG.

PMID:42682657 | PMC:PMC13530114 | DOI:10.1007/s13167-026-00465-4

  •  
  • ↗️
↗️

Solving Combinatorial Counting Problems with Weighted First-Order Model Counting

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yuanhong Wang · Juhua Pu · Yuxu Zhou · Yuyi Wang · Ond\v{r}ej Ku\v{z}elka
arXiv:2605.24845v1 Announce Type: new Abstract: Combinatorial counting problems pervade artificial intelligence, statistics, and discrete mathematics. Whether the task is enumerating subsets, multisets, permutations, partitions, or compositions under structural and arithmetic constraints, solving it remains a stubbornly manual exercise. Closed-form derivations are powerful but brittle, while naive encodings to propositional model counting or constraint satisfaction destroy the exchangeability that makes counting tractable in the first place. We present Cofola (COmbinatorial counting LAnguage with First-Order logic), a typed declarative language whose primitives are the combinatorial objects that recur in everyday counting questions, including sets, bags, tuples, sequences, circles, partitions, and compositions, together with natural relational and arithmetic constraints over them. A denotational semantics maps every Cofola program to a well-defined combinatorial counting problem, and a three-phase compilation pipeline (preprocessing, decomposition, and symmetry-preserving encoding) reduces this problem to a weighted first-order model counting (WFOMC) instance augmented with coefficient-extraction constraints. To stay inside known domain-liftable fragments whenever possible, the encoding groups indistinguishable entities, breaks the symmetry of unordered groupings lexicographically, and encodes sequences and circles via order axioms. On a suite of representative combinatorial counting problems, ranging from textbook math problems to multi-object scenarios that the closest prior framework cannot express, Cofola produces concise specifications and a uniform solving pipeline that is practical end-to-end.
  •  
  • ↗️
↗️

Liver-specific <i>SIRT1</i> knockout-induced hyperglycemia promotes spontaneous lung adenocarcinomas through HSF1-MDM2

✇Oncogene - Issue - nature.com science feeds
By: Yiwei Cao · Jingyi Wei · Qiang Chen · Haitao Wang · Jiao Mu · Shuhan Lu · Menghui Yuan · Mohammed A. El-Magd · Bichan Xu · Ruihong Wang · Peng Yuan

Oncogene, Published online: 24 May 2026; doi:10.1038/s41388-026-03826-5

Liver-specific SIRT1 knockout-induced hyperglycemia promotes spontaneous lung adenocarcinomas through HSF1-MDM2
  •  
  • ↗️
↗️

Multi-omics analysis of glutamine and fish collagen peptides in alleviating post-antibiotic Streptococcus pneumoniae injury in feline lung cells

✇Omics In Lung
By: Huasong Bai · Tong Liu · Hengyan Wang · Hongchen Jin · Yunliang Li · Zhanzhong Wang

Exp Ther Med. 2026 Mar 30;31(6):148. doi: 10.3892/etm.2026.13143. eCollection 2026 Jun.

ABSTRACT

Streptococcus pneumoniae (SP) infection often leads to persistent lung injury even after antibiotic treatment. Despite this phenomenon, the mechanisms underlying host cell recovery remain poorly understood. Upon breaching the epithelial barrier, SP primarily targets the pulmonary interstitial cells, which constitute the major mesenchymal component of the lung. These cells serve as essential effectors of tissue repair, extracellular matrix remodeling and epithelial restoration. Therefore, a feline pulmonary interstitial cell (FCA-L2) model of SP infection was established to investigate the protective effects of glutamine (GLU) and fish collagen peptides (FCP) through integrated transcriptomic and metabolomic analyses. Cells were infected with SP (0.05 McFarland units for 4 h) and then treated with doxycycline (7.5 µg/ml for 18 h) followed by GLU (40 mM) or FCP (500 µg/ml). Notably, SP infection increased lactate dehydrogenase (LDH) release by 3.5-fold, induced secretion of IL-1β, TNF-α and IL-8, disrupted tight-junction proteins (claudin, ZO-1 and occludin) and caused oxidative imbalance and apoptosis despite antibiotic (doxycycline) treatment. However, treatment with GLU or FCP significantly reduced LDH release by ~40%, restored junctional proteins, suppressed inflammatory cytokines and enhanced antioxidant enzyme activities. Multi-omics analysis revealed that GLU promoted amino acid biosynthesis and energy metabolism and suppressed aminoacyl-tRNA synthetases and cell-cycle regulators, thereby enhancing metabolic adaptability. By contrast, FCP activated amino and nucleotide sugar metabolism, increased polyunsaturated fatty-acid synthesis and supported glycocalyx repair and membrane reconstruction. GLU and FCP provided complementary metabolic and structural protection, which mitigated post-infectious stress and promoted cellular recovery. The findings of the present study underscore the potential of bioactive food-derived compounds as adjunctive therapies that may accelerate lung tissue repair and enhance the efficacy of conventional antibiotics.

PMID:41988354 | PMC:PMC13077270 | DOI:10.3892/etm.2026.13143

  •  
  • ↗️
↗️

Pan-neurodegeneration proteomics reveals disease subtypes and molecular signatures

✇Cell
By: Him K. Shrestha · Huan Sun · Jay M. Yarbro · DongGeun Lee · Danting Liu · Erming Wang · Meghan McReynolds · Nan Zhang · Boer Xie · Shu Yang · Kaiwen Yu · Suresh Poudel · Yuxin Li · Zuo-Fei Yuan · Dehui Kong · Minghui Wang · Zhen Wang · Mingming Niu · Hong Wang · Masihuz Zaman · Ju Wang · David R. Vanderwall · Yu Sun · Zhiping Wu · Ping-Chung Chen · Bing Bai · Anthony A. High · Júlia Faura · Chunyu Liu · David A. Bennett · Erik C.B. Johnson · Nicholas T. Seyfried · Allan I. Levey · Vahram Haroutunian · Geidy E. Serrano · Thomas G. Beach · Michael DeTure · Takahisa Kanekiyo · Ronald C. Petersen · Guojun Bu · Pamela J. McLean · Dennis W. Dickson · Rosa Rademakers · Gang Yu · Xusheng Wang · Bin Zhang · Junmin Peng
A pan-neurodegeneration atlas built from multilayer, deep proteomics of 2,279 brain samples across 6 major diseases integrates whole proteome, detergent-insoluble proteome, and posttranslational modifications to enable intra- and inter-disease comparisons to reveal disease-specific subtypes and dysregulated pathways, while identifying shared changes such as GPNMB upregulation and NPTX2 downregulation.
  •  
  • ↗️
↗️

Vero: An Open RL Recipe for General Visual Reasoning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Gabriel Sarch · Linrong Cai · Qunzhong Wang · Haoyang Wu · Danqi Chen · Zhuang Liu
arXiv:2604.04917v2 Announce Type: cross Abstract: What does it take to build a visual reasoner that works across charts, science, spatial understanding, and open-ended tasks? The strongest vision-language models (VLMs) show such broad visual reasoning is within reach, but the recipe behind them remains unclear, locked behind proprietary reinforcement learning (RL) pipelines with non-public data. We introduce Vero, a family of fully open VLMs that matches or exceeds existing open-weight models across diverse visual reasoning tasks. We scale RL data and rewards across six broad task categories, constructing Vero-600K, a 600K-sample dataset from 59 datasets, and designing task-routed rewards that handle heterogeneous answer formats. Vero achieves state-of-the-art performance, improving over four base models by 3.6-5.3 points on average across VeroEval, our suite of 30 challenging benchmarks. Starting from Qwen3-VL-8B-Instruct, Vero outperforms Qwen3-VL-8B-Thinking on 23 of 30 benchmarks without additional proprietary thinking data. When trained from the same base model, Vero-600K exceeds existing RL datasets across task categories. Systematic ablations reveal that different task categories elicit qualitatively distinct reasoning patterns that transfer poorly in isolation, suggesting that broad data coverage is the primary driver of strong RL scaling. All data, code, and models are released.
  •  
  • ↗️
↗️

UniMark: Artificial Intelligence Generated Content Identification Toolkit

✇cs.AI, q-bio.NC updates on arXiv.org
By: Meilin Li · Ji He · Yi Yu · Jia Xu · Shanzhe Lei · Yan Teng · Yingchun Wang · Xuhong Wang
arXiv:2512.12324v3 Announce Type: replace-cross Abstract: The rapid proliferation of Artificial Intelligence Generated Content has precipitated a crisis of trust and urgent regulatory demands. However, existing identification tools suffer from fragmentation and a lack of support for visible compliance marking. To address these gaps, we introduce the \textbf{UniMark}, an open-source, unified framework for multimodal content governance. Our system features a modular unified engine that abstracts complexities across text, image, audio, and video modalities. Crucially, we propose a novel dual-operation strategy, natively supporting both \emph{Hidden Watermarking} for copyright protection and \emph{Visible Marking} for regulatory compliance. Furthermore, we establish a standardized evaluation framework with three specialized benchmarks (Image/Video/Audio-Bench) to ensure rigorous performance assessment. This toolkit bridges the gap between advanced algorithms and engineering implementation, fostering a more transparent and secure digital ecosystem.
  •  
  • ↗️
↗️

Semantic Refinement with LLMs for Graph Representations

✇cs.AI, q-bio.NC updates on arXiv.org
By: Safal Thapaliya · Zehong Wang · Jiazheng Li · Ziming Li · Yanfang Ye · Chuxu Zhang
arXiv:2512.21106v2 Announce Type: replace-cross Abstract: Graph-structured data exhibit substantial heterogeneity in where their predictive signals originate: in some domains, node-level semantics dominate, while in others, structural patterns play a central role. This structure-semantics heterogeneity implies that no graph learning model with a fixed inductive bias can generalize optimally across diverse graph domains. However, most existing methods address this challenge from the model side by incrementally injecting new inductive biases, which remains fundamentally limited given the open-ended diversity of real-world graphs. In this work, we take a data-centric perspective and treat node semantics as a task-adaptive variable. We propose a Graph-Exemplar-guided Semantic Refinement (GES) framework for graph representation learning which -- unlike existing LLM-enhanced methods that generate node descriptions without graph context -- leverages structurally and semantically similar nodes from the graph itself to guide semantic refinement. Specifically, a GNN is first trained to produce predictive states, which along with structural and semantic similarity are used to retrieve in-graph exemplars that inform an LLM in refining node descriptions. We evaluate our approach on both text-rich and text-free graphs. Results show consistent improvements on semantics-rich and structure-dominated graphs, demonstrating the effectiveness of data-centric semantic refinement under structure-semantics heterogeneity.
  •  
  • ↗️
↗️

Improving Ensemble Forecasts of Abnormally Deflecting Tropical Cyclones with Fused Atmosphere-Ocean-Terrain Data

✇cs.AI, q-bio.NC updates on arXiv.org
By: Qixiang Li · Yuan Zhou · Shuwei Huo · Chong Wang · Xiaofeng Li
arXiv:2603.29200v2 Announce Type: replace-cross Abstract: Deep learning-based tropical cyclone (TC) forecasting methods have demonstrated significant potential and application advantages, as they feature much lower computational cost and faster operation speed than numerical weather prediction models. However, existing deep learning methods still have key limitations: they can only process a single type of sequential trajectory data or homogeneous meteorological variables, and fail to achieve accurate forecasting of abnormal deflected TCs. To address these challenges, we present two groundbreaking contributions. First, we have constructed a multimodal and multi-source dataset named AOT-TCs for TC forecasting in the Northwest Pacific basin. As the first dataset of its kind, it innovatively integrates heterogeneous variables from the atmosphere, ocean, and land, thus obtaining a comprehensive and information-rich meteorological dataset. Second, based on the AOT-TCs dataset, we propose a forecasting model that can handle both normal and abnormally deflected TCs. This is the first TC forecasting model to adopt an explicit atmosphere-ocean-terrain coupling architecture, enabling it to effectively capture complex interactions across physical domains. Extensive experiments on all TC cases in the Northwest Pacific from 2017 to 2024 show that our model achieves state-of-the-art performance in TC forecasting: it not only significantly improves the forecasting accuracy of normal TCs but also breaks through the technical bottleneck in forecasting abnormally deflected TCs.
  •  
  • ↗️
↗️

Improving Ensemble Forecasts of Abnormally Deflecting Tropical Cyclones with Fused Atmosphere-Ocean-Terrain Data

✇cs.AI, q-bio.NC updates on arXiv.org
By: Qixiang Li · Shuwei Huo · Chong Wang · Xiaofeng Li · Yuan Zhou
arXiv:2603.29200v1 Announce Type: cross Abstract: Deep learning-based tropical cyclone (TC) forecasting methods have demonstrated significant potential and application advantages, as they feature much lower computational cost and faster operation speed than numerical weather prediction models. However, existing deep learning methods still have key limitations: they can only process a single type of sequential trajectory data or homogeneous meteorological variables, and fail to achieve accurate forecasting of abnormal deflected TCs. To address these challenges, we present two groundbreaking contributions. First, we have constructed a multimodal and multi-source dataset named AOT-TCs for TC forecasting in the Northwest Pacific basin. As the first dataset of its kind, it innovatively integrates heterogeneous variables from the atmosphere, ocean, and land, thus obtaining a comprehensive and information-rich meteorological dataset. Second, based on the AOT-TCs dataset, we propose a forecasting model that can handle both normal and abnormally deflected TCs. This is the first TC forecasting model to adopt an explicit atmosphere-ocean-terrain coupling architecture, enabling it to effectively capture complex interactions across physical domains. Extensive experiments on all TC cases in the Northwest Pacific from 2017 to 2024 show that our model achieves state-of-the-art performance in TC forecasting: it not only significantly improves the forecasting accuracy of normal TCs but also breaks through the technical bottleneck in forecasting abnormally deflected TCs.
  •  
  • ↗️
↗️

IMAGAgent: Orchestrating Multi-Turn Image Editing via Constraint-Aware Planning and Reflection

✇cs.AI, q-bio.NC updates on arXiv.org
By: Fei Shen · Chengyu Xie · Lihong Wang · Zhanyi Zhang · Xin Jiang · Xiaoyu Du · Jinhui Tang
arXiv:2603.29602v1 Announce Type: cross Abstract: Existing multi-turn image editing paradigms are often confined to isolated single-step execution. Due to a lack of context-awareness and closed-loop feedback mechanisms, they are prone to error accumulation and semantic drift during multi-turn interactions, ultimately resulting in severe structural distortion of the generated images. For that, we propose \textbf{IMAGAgent}, a multi-turn image editing agent framework based on a "plan-execute-reflect" closed-loop mechanism that achieves deep synergy among instruction parsing, tool scheduling, and adaptive correction within a unified pipeline. Specifically, we first present a constraint-aware planning module that leverages a vision-language model (VLM) to precisely decompose complex natural language instructions into a series of executable sub-tasks, governed by target singularity, semantic atomicity, and visual perceptibility. Then, the tool-chain orchestration module dynamically constructs execution paths based on the current image, the current sub-task, and the historical context, enabling adaptive scheduling and collaborative operation among heterogeneous operation models covering image retrieval, segmentation, detection, and editing. Finally, we devise a multi-expert collaborative reflection mechanism where a central large language model (LLM) receives the image to be edited and synthesizes VLM critiques into holistic feedback, simultaneously triggering fine-grained self-correction and recording feedback outcomes to optimize future decisions. Extensive experiments on our constructed \textbf{MTEditBench} and the MagicBrush dataset demonstrate that IMAGAgent achieves performance significantly superior to existing methods in terms of instruction consistency, editing precision, and overall quality. The code is available at https://github.com/hackermmzz/IMAGAgent.git.
  •  
  • ↗️
↗️

FERA: A Pose-Based Framework for Rule-Grounded Multimedia Decision Support with a Foil Fencing Case Study

✇cs.AI, q-bio.NC updates on arXiv.org
By: Ziwen Chen · Zhong Wang
arXiv:2509.18527v5 Announce Type: replace Abstract: Multimedia decision support requires more than recognition; it requires explicit state estimates that can be checked against rules, audited by humans, and consumed by downstream decision logic. We present the FEncing Referee Assistant (FERA), a pose-based framework for this setting, and study it through foil fencing, where decisions depend on fast bilateral motion and right-of-way rules. The framework separates canonical participant tracking, kinematic tokenization, calibrated temporal perception, a compact structured decision layer, and an explanation-oriented retrieval interface. We also release an audited benchmark with adjudicated labels and fixed folds for reproducible evaluation. Under a shared protocol, a lightweight lifted-depth sidecar strengthens the best graph-based perception model, while a compact structured classifier on the fixed two-dimensional token stream reaches 0.624 accuracy and a 0.632 macro-averaged F1 score on the final Left / Right / None decision. The case study supports a broader design lesson: keep the boundary between perception and rule application explicit, preserve uncertainty, and choose the perception front end according to the downstream operating point.
  •  
  • ↗️
↗️

Extracellular matrix sensing regulates intratumoral heterogeneity of autophagic flux

✇Cell
By: Mohamad Assi · Ruohong Wang · Emily A. Kawaler · Albert S.W. Sohn · M. Zahidunnabi Dewan · Despoina Kalfakakou · Joel Encarnacion-Rosado · Kevin S. Kapner · Koelina Ganguly · Joao A. Paulo · Diane M. Simeone · Andrew J. Aguirre · Robert S. Banh · Alec C. Kimmelman
Human pancreatic cancer cells sense specific components of the ECM to fine-tune their autophagy flux levels, conferring the ability to coordinate proliferation, survival, and responsiveness to chemotherapy. Targeting ECM sensing may turn this ability into a therapeutic opportunity.
  •  
  • ↗️
↗️

LatentChem: From Textual CoT to Latent Thinking in Chemical Reasoning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xinwu Ye · Yicheng Mao · Jia Zhang · Yimeng Liu · Li Hao · Fang Wu · Zhiwei Li · Yuxuan Liao · Zehong Wang · Yingcheng Wu · Zhiyuan Liu · Zhenfei Yin · Li Yuan · Philip Torr · Huan Sun · Xiangxiang Zeng · Mengdi Wang · Le Cong · Shenghua Gao · Xiangru Tang
arXiv:2602.07075v4 Announce Type: replace-cross Abstract: Chemical large language models (LLMs) predominantly rely on explicit Chain-of-Thought (CoT) in natural language to perform complex reasoning. However, chemical reasoning is inherently continuous and structural, and forcing it into discrete linguistic tokens introduces a fundamental representation mismatch that constrains both efficiency and performance. We introduce LatentChem, a latent reasoning interface that decouples chemical computation from textual generation, enabling models to perform multi-step reasoning directly in continuous latent space while emitting language only for final outputs. Remarkably, we observe a consistent emergent behavior: when optimized solely for task success, models spontaneously internalize reasoning, progressively abandoning verbose textual derivations in favor of implicit latent computation. This shift is not merely stylistic but computationally advantageous. Across diverse chemical reasoning benchmarks, LatentChem achieves a 59.88\% non-tie win rate over strong CoT-based baselines on ChemCoTBench, while delivering a 10.84$\times$ average reduction in reasoning overhead. Our results provide empirical evidence that chemical reasoning is more naturally and effectively realized as continuous latent dynamics rather than discretized linguistic trajectories.
  •  
  • ↗️
↗️

Towards Personalized Deep Research: Benchmarks and Evaluations

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yuan Liang · Jiaxian Li · Yuqing Wang · Piaohong Wang · Motong Tian · Pai Liu · Shuofei Qiao · Runnan Fang · He Zhu · Ge Zhang · Minghao Liu · Yuchen Eleanor Jiang · Ningyu Zhang · Wangchunshu Zhou
arXiv:2509.25106v3 Announce Type: replace-cross Abstract: Deep Research Agents (DRAs) can autonomously conduct complex investigations and generate comprehensive reports, demonstrating strong real-world potential. However, existing evaluations mostly rely on close-ended benchmarks, while open-ended deep research benchmarks remain scarce and typically neglect personalized scenarios. To bridge this gap, we introduce Personalized Deep Research Bench (PDR-Bench), the first benchmark for evaluating personalization in DRAs. It pairs 50 diverse research tasks across 10 domains with 25 authentic user profiles that combine structured persona attributes with dynamic real-world contexts, yielding 250 realistic user-task queries. To assess system performance, we propose the PQR Evaluation Framework, which jointly measures Personalization Alignment, Content Quality, and Factual Reliability. Our experiments on a range of systems highlight current capabilities and limitations in handling personalized deep research. This work establishes a rigorous foundation for developing and evaluating the next generation of truly personalized AI research assistants.
  •  
  • ↗️
↗️

Dynamic Adversarial Reinforcement Learning for Robust Multimodal Large Language Models

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yicheng Bao · Xuhong Wang · Qiaosheng Zhang · Chaochao Lu · Xia Hu · Xin Tan
arXiv:2602.22227v3 Announce Type: replace-cross Abstract: Despite their impressive capabilities, Multimodal Large Language Models (MLLMs) exhibit perceptual fragility when confronted with visually complex scenes. This weakness stems from a reliance on finite training datasets, which are prohibitively expensive to scale and impose a ceiling on model robustness. We introduce \textbf{AOT-SFT}, a large-scale adversarial dataset for bootstrapping MLLM robustness. Building on this, we propose \textbf{AOT (Adversarial Opponent Training)}, a self-play framework that forges MLLM robustness by creating its own training data. Our method orchestrates a co-evolution between an image-editing Attacker and a Defender MLLM, where the Attacker generates a diverse and dynamic curriculum of image manipulations, forcing the Defender to adapt and improve. Extensive experiments demonstrate that AOT enhances the Defender's perceptual robustness and reduces hallucinations, establishing a scalable paradigm for training more reliable MLLMs.
  •  
  • ↗️
↗️

Rubrics as an Attack Surface: Stealthy Preference Drift in LLM Judges

✇cs.AI, q-bio.NC updates on arXiv.org
By: Ruomeng Ding · Yifei Pang · He Sun · Yizhong Wang · Zhiwei Steven Wu · Zhun Deng
arXiv:2602.13576v1 Announce Type: cross Abstract: Evaluation and alignment pipelines for large language models increasingly rely on LLM-based judges, whose behavior is guided by natural-language rubrics and validated on benchmarks. We identify a previously under-recognized vulnerability in this workflow, which we term Rubric-Induced Preference Drift (RIPD). Even when rubric edits pass benchmark validation, they can still produce systematic and directional shifts in a judge's preferences on target domains. Because rubrics serve as a high-level decision interface, such drift can emerge from seemingly natural, criterion-preserving edits and remain difficult to detect through aggregate benchmark metrics or limited spot-checking. We further show this vulnerability can be exploited through rubric-based preference attacks, in which benchmark-compliant rubric edits steer judgments away from a fixed human or trusted reference on target domains, systematically inducing RIPD and reducing target-domain accuracy up to 9.5% (helpfulness) and 27.9% (harmlessness). When these judgments are used to generate preference labels for downstream post-training, the induced bias propagates through alignment pipelines and becomes internalized in trained policies. This leads to persistent and systematic drift in model behavior. Overall, our findings highlight evaluation rubrics as a sensitive and manipulable control interface, revealing a system-level alignment risk that extends beyond evaluator reliability alone. The code is available at: https://github.com/ZDCSlab/Rubrics-as-an-Attack-Surface. Warning: Certain sections may contain potentially harmful content that may not be appropriate for all readers.
  •  
  • ↗️
↗️

OPBench: A Graph Benchmark to Combat the Opioid Crisis

✇cs.AI, q-bio.NC updates on arXiv.org
By: Tianyi Ma · Yiyang Li · Yiyue Qian · Zheyuan Zhang · Zehong Wang · Chuxu Zhang · Yanfang Ye
arXiv:2602.14602v1 Announce Type: cross Abstract: The opioid epidemic continues to ravage communities worldwide, straining healthcare systems, disrupting families, and demanding urgent computational solutions. To combat this lethal opioid crisis, graph learning methods have emerged as a promising paradigm for modeling complex drug-related phenomena. However, a significant gap remains: there is no comprehensive benchmark for systematically evaluating these methods across real-world opioid crisis scenarios. To bridge this gap, we introduce OPBench, the first comprehensive opioid benchmark comprising five datasets across three critical application domains: opioid overdose detection from healthcare claims, illicit drug trafficking detection from digital platforms, and drug misuse prediction from dietary patterns. Specifically, OPBench incorporates diverse graph structures, including heterogeneous graphs and hypergraphs, to preserve the rich and complex relational information among drug-related data. To address data scarcity, we collaborate with domain experts and authoritative institutions to curate and annotate datasets while adhering to privacy and ethical guidelines. Furthermore, we establish a unified evaluation framework with standardized protocols, predefined data splits, and reproducible baselines to facilitate fair and systematic comparison among graph learning methods. Through extensive experiments, we analyze the strengths and limitations of existing graph learning methods, thereby providing actionable insights for future research in combating the opioid crisis. Our source code and datasets are available at https://github.com/Tianyi-Billy-Ma/OPBench.
  •  
  • ↗️
❌