arXiv subjects

Wenxuan Wang

Publications and source records attributed to Wenxuan Wang.

At least 55 records · Page 3Linked to original sources

arXiv2026

RefereeBench: Are Video MLLMs Ready to be Multi-Sport Referees

While Multimodal Large Language Models (MLLMs) excel at generic video understanding, their ability to support specialized, rule-grounded decision-making remains insufficiently explored. In this paper, we introduce RefereeBench, the first large-scale benchmark for evaluating MLLMs as automatic sports referees. Spanning 11 sports with 925 curated videos and 6,475 QA pairs, RefereeBench evaluates five core officiating abilities: foul existence, foul and penalty classification, foul and penalty reasoning, entity perception, and temporal grounding. The benchmark is fully human-annotated to ensure high-quality annotations grounded in authentic officiating logic and multimodal evidence. Extensive evaluations of state-of-the-art MLLMs show that even the strongest models, such as Doubao-Seed-1.8 and Gemini-3-Pro, achieve only around 60% accuracy, while the strongest open-source model, Qwen3-VL, reaches only 47%. These results indicate that current models remain far from being reliable sports referees. Further analysis shows that while models can often identify incidents and involved entities, they struggle with rule application and temporal grounding, and frequently over-call fouls on normal clips. Our benchmark highlights the need for future MLLMs that better integrate domain knowledge and multimodal understanding, advancing trustworthy AI-assisted officiating and broader multimodal decision-making.

cs.CV↗

arXiv2026

SECURE: Stable Early Collision Understanding via Robust Embeddings in Autonomous Driving

While deep learning has significantly advanced accident anticipation, the robustness of these safety-critical systems against real-world perturbations remains a major challenge. We reveal that state-of-the-art models like CRASH, despite their high performance, exhibit significant instability in predictions and latent representations when faced with minor input perturbations, posing serious reliability risks. To address this, we introduce SECURE - Stable Early Collision Understanding Robust Embeddings, a framework that formally defines and enforces model robustness. SECURE is founded on four key attributes: consistency and stability in both prediction space and latent feature space. We propose a principled training methodology that fine-tunes a baseline model using a multi-objective loss, which minimizes divergence from a reference model and penalizes sensitivity to adversarial perturbations. Experiments on DAD and CCD datasets demonstrate that our approach not only significantly enhances robustness against various perturbations but also improves performance on clean data, achieving new state-of-the-art results.

cs.LG↗

arXiv2026

Privacy-Preserving Formation Control for Networked Underactuated USVs: A Passivity-Based Approach

This paper studies coordinated trajectory planning and tracking control for multiple unmanned surface vessels (USVs) under strict privacy requirements. To avoid the privacy risks associated with direct position sharing in conventional cooperative methods, the proposed approach adopts an estimated fleet centroid as the only shared variable, preventing individual trajectory disclosure while enabling coordination. Based on this interaction mechanism, a formation-oriented trajectory is generated for the fleet. The collective dynamics are modeled using Port-Hamiltonian systems, and a passivity-based tracking controller is designed for each USV to accurately follow the planned trajectories. The stability of the closed-loop system is rigorously proven, and experiments on a real USV platform confirm effective formation tracking and privacy preservation. The proposed result extends and validates through experimental results the approach in [26] that was limited to idealized pointmass models and lacked a feedback control.

math.OC↗

arXiv2026

Toward Personalized LLM-Powered Agents: Foundations, Evaluation, and Future Directions

Large language models have enabled agentic systems that reason, plan, and interact with tools and environments to accomplish complex tasks. As these agents operate over extended interaction horizons, their effectiveness increasingly depends on adapting behavior to individual users and maintaining continuity across interactions, giving rise to personalized LLM-powered agents (PLAs). In such long-term, user-dependent settings, personalization permeates the entire decision pipeline rather than remaining confined to surface-level response generation. This survey provides a capability-oriented review of personalized LLM-powered agents. Existing work is organized around four interdependent capabilities: profile modeling, memory, planning, and action execution. Using this taxonomy, representative methods are synthesized and analyzed to illustrate how user signals are represented, propagated, and utilized across the agent pipeline, highlighting cross-component interactions and recurring design challenges. Evaluation metrics and benchmarking paradigms tailored to personalized agents are further examined, along with application scenarios ranging from conversational assistants to domain-specific expert systems. By clarifying the design space of personalization in agent systems, this survey provides a structured foundation for developing more user-aligned, adaptive, and deployable LLM-powered agents.

cs.AI↗

arXiv2026

ST-VLA: Enabling 4D-Aware Spatiotemporal Understanding for General Robot Manipulation

Robotic manipulation in open-world environments requires reasoning across semantics, geometry, and long-horizon action dynamics. Existing hierarchical Vision-Language-Action (VLA) frameworks typically use 2D representations to connect high-level reasoning with low-level control, but lack depth awareness and temporal consistency, limiting robustness in complex 3D scenes. We propose ST-VLA, a hierarchical VLA framework using a unified 3D-4D representation to bridge perception and action. ST-VLA converts 2D guidance into 3D trajectories and generates smooth spatial masks that capture 4D spatio-temporal context, providing a stable interface between semantic reasoning and continuous control. To enable effective learning of such representations, we introduce ST-Human, a large-scale human manipulation dataset with 14 tasks and 300k episodes, annotated with 2D, 3D, and 4D supervision via a semi-automated pipeline. Using ST-Human, we train ST-VLM, a spatio-temporal vision-language model that generates spatially grounded and temporally coherent 3D representations to guide policy execution. The smooth spatial masks focus on task-relevant geometry and stabilize latent representations, enabling online replanning and long-horizon reasoning. Experiments on RLBench and real-world manipulation tasks show that \method significantly outperforms state-of-the-art baselines, improving zero-shot success rates by 44.6% and 30.3%. These results demonstrate that offloading spatio-temporal reasoning to VLMs with unified 3D-4D representations substantially improves robustness and generalization for open-world robotic manipulation. Project website: https://oucx117.github.io/ST-VLA/.

cs.RO↗

arXiv2026

Benchmarking MLLM-based Web Understanding: Reasoning, Robustness and Safety

Multimodal large language models (MLLMs) are increasingly deployed as the core reasoning engine for web-facing systems, powering GUI agents and front-end automation that must interpret page structure, select actionable widgets, and execute multi-step interactions reliably. However, existing benchmarks largely emphasize visual perception or UI code generation, showing insufficient evaluation on the reasoning, robustness and safety capability required for end-to-end web applications. To bridge the gap, we introduce a comprehensive web understanding benchmark, named WebRRSBench, that jointly evaluates Reasoning, Robustness, and Safety across eight tasks, such as position relationship reasoning, color robustness, and safety critical detection, etc. The benchmark is constructed from 729 websites and contains 3799 QA pairs that probe multi-step inference over page structure, text, widgets, and safety-critical interactions. To ensure reliable measurement, we adopt standardized prompts, a protocolized and deterministic evaluation pipeline, and multi-stage quality control combining automatic checks with targeted human verification. We evaluate 11 MLLMs on WebRRSBench. The results reveal significant gaps: models still struggle with compositional and cross-element reasoning over realistic layouts, show limited robustness when facing perturbations in user interfaces and content such as layout rearrangements or visual style shifts, and are rather conservative in recognizing and avoiding safety critical or irreversible actions. Our code and appendix are available at https: //github.com/annoy-worker/WebRSSBench.

cs.AI↗

arXiv2026

Order Is Not Layout: Order-to-Space Bias in Image Generation

We study a systematic bias in modern image generation models: the mention order of entities in text spuriously determines spatial layout and entity--role binding. We term this phenomenon Order-to-Space Bias (OTS) and show that it arises in both text-to-image and image-to-image generation, often overriding grounded cues and causing incorrect layouts or swapped assignments. To quantify OTS, we introduce OTS-Bench, which isolates order effects with paired prompts differing only in entity order and evaluates models along two dimensions: homogenization and correctness. Experiments show that Order-to-Space Bias (OTS) is widespread in modern image generation models, and provide evidence that it is primarily data-driven and manifests during the early stages of layout formation. Motivated by this insight, we show that both targeted fine-tuning and early-stage intervention strategies can substantially reduce OTS, while preserving generation quality.

cs.CL↗

arXiv2026

TrustMH-Bench: A Comprehensive Benchmark for Evaluating the Trustworthiness of Large Language Models in Mental Health

While Large Language Models (LLMs) demonstrate significant potential in providing accessible mental health support, their practical deployment raises critical trustworthiness concerns due to the domains high-stakes and safety-sensitive nature. Existing evaluation paradigms for general-purpose LLMs fail to capture mental health-specific requirements, highlighting an urgent need to prioritize and enhance their trustworthiness. To address this, we propose TrustMH-Bench, a holistic framework designed to systematically quantify the trustworthiness of mental health LLMs. By establishing a deep mapping from domain-specific norms to quantitative evaluation metrics, TrustMH-Bench evaluates models across eight core pillars: Reliability, Crisis Identification and Escalation, Safety, Fairness, Privacy, Robustness, Anti-sycophancy, and Ethics. We conduct extensive experiments across six general-purpose LLMs and six specialized mental health models. Experimental results indicate that the evaluated models underperform across various trustworthiness dimensions in mental health scenarios, revealing significant deficiencies. Notably, even generally powerful models (e.g., GPT-5.1) fail to maintain consistently high performance across all dimensions. Consequently, systematically improving the trustworthiness of LLMs has become a critical task. Our data and code are released.

cs.CL↗

arXiv2026

Interaction2Code: Benchmarking MLLM-based Interactive Webpage Code Generation from Interactive Prototyping

Multimodal Large Language Models (MLLMs) have demonstrated remarkable performance on the design-to-code task, i.e., generating UI code from UI mock-ups. However, existing benchmarks only contain static web pages for evaluation and ignore the dynamic interaction, limiting the practicality, usability and user engagement of the generated webpages. To bridge these gaps, we present the first systematic investigation of MLLMs in generating interactive webpages. Specifically, we formulate the Interaction-to-Code task and establish the Interaction2Code benchmark, encompassing 127 unique webpages and 374 distinct interactions across 15 webpage types and 31 interaction categories. Through comprehensive experiments utilizing state-of-the-art (SOTA) MLLMs, evaluated via both automatic metrics and human assessments, we identify four critical limitations of MLLM on Interaction-to-Code task: (1) inadequate generation of interaction compared with full page, (2) prone to ten types of failure, (3) poor performance on visually subtle interactions, and (4) insufficient undestanding on interaction when limited to single-modality visual descriptions. To address these limitations, we propose four enhancement strategies: Interactive Element Highlighting, Failureaware Prompting (FAP), Visual Saliency Enhancement, and Visual-Textual Descriptions Combination, all aiming at improving MLLMs' performance on the Interaction-toCode task. The Interaction2Code benchmark and code are available in https://github. com/WebPAI/Interaction2Code.

cs.SE↗

arXiv2026

EfficientPosterGen: Semantic-aware Efficient Poster Generation via Token Compression and Accurate Violation Detection

Automated academic poster generation aims to distill lengthy research papers into concise, visually coherent presentations. Existing Multimodal Large Language Models (MLLMs) based approaches, however, suffer from three critical limitations: low information density in full-paper inputs, excessive token consumption, and unreliable layout verification. We present EfficientPosterGen, an end-to-end framework that addresses these challenges through semantic-aware retrieval and token-efficient multimodal generation. EfficientPosterGen introduces three core innovations: (1) Semantic-aware Key Information Retrieval (SKIR), which constructs a semantic contribution graph to model inter-segment relationships and selectively preserves important content; (2) Visual-based Context Compression (VCC), which renders selected text segments into images to shift textual information into the visual modality, significantly reducing token usage while generating poster-ready bullet points; and (3) Agentless Layout Violation Detection (ALVD), a deterministic color-gradient-based algorithm that reliably detects content overflow and spatial sparsity without auxiliary MLLMs. Extensive experiments demonstrate that EfficientPosterGen achieves substantial improvements in token efficiency and layout reliability while maintaining high poster quality, offering a scalable solution for automated academic poster generation. Our code is available at https://github.com/vinsontang1/EfficientPosterGen-Code.

cs.CV↗

arXiv2026

Scaling World Model for Hierarchical Manipulation Policies

Vision-Language-Action (VLA) models are promising for generalist robot manipulation but remain brittle in out-of-distribution (OOD) settings, especially with limited real-robot data. To resolve the generalization bottleneck, we introduce a hierarchical Vision-Language-Action framework \our{} that leverages the generalization of large-scale pre-trained world model for robust and generalizable VIsual Subgoal TAsk decomposition VISTA. Our hierarchical framework \our{} consists of a world model as the high-level planner and a VLA as the low-level executor. The high-level world model first divides manipulation tasks into subtask sequences with goal images, and the low-level policy follows the textual and visual guidance to generate action sequences. Compared to raw textual goal specification, these synthesized goal images provide visually and physically grounded details for low-level policies, making it feasible to generalize across unseen objects and novel scenarios. We validate both visual goal synthesis and our hierarchical VLA policies in massive out-of-distribution scenarios, and the performance of the same-structured VLA in novel scenarios could boost from 14% to 69% with the guidance generated by the world model. Results demonstrate that our method outperforms previous baselines with a clear margin, particularly in out-of-distribution scenarios. Project page: \href{https://vista-wm.github.io/}{https://vista-wm.github.io}

cs.RO↗

arXiv2026

Understanding and Bridging the Planner-Coder Gap: A Systematic Study on the Robustness of Multi-Agent Systems for Code Generation

Multi-agent systems (MASs) have emerged as a promising paradigm for automated code generation, demonstrating impressive performance on established benchmarks. Despite their prosperous development, the fundamental mechanisms underlying their robustness remain poorly understood, raising critical concerns for real-world deployment. This paper conducts a systematic empirical study to uncover the internal robustness flaws of MASs using a mutation-based methodology. By designing a testing pipeline incorporating semantic-preserving mutation operators and a novel fitness function, we assess mainstream MASs across multiple datasets and LLMs. Our findings reveal substantial robustness flaws: semantically equivalent inputs cause drastic performance drops, with MASs failing to solve 7.9\%--83.3\% of problems they initially resolved successfully. Through comprehensive failure analysis, we discover a fundamental cause underlying these robustness issues: the \textit{planner-coder gap}, which accounts for 75.3\% of failures. This gap arises from information loss in the multi-stage transformation process where planning agents decompose requirements into underspecified plans, and coding agents subsequently misinterpret intricate logic during code generation. Based on this formulated information transformation process, we propose a \textit{repairing method} that mitigates information loss through multi-prompt generation and introduces a monitor agent to bridge the planner-coder gap. Evaluation shows that our repairing method effectively enhances the robustness of MASs by solving 40.0\%--88.9\% of identified failures. Our work uncovers critical robustness flaws in MASs and provides effective mitigation strategies, contributing essential insights for developing more reliable MASs for code generation.

cs.SE↗

arXiv2026

On the Failure of Latent State Persistence in Large Language Models

While Large Language Models (LLMs) excel in reasoning, whether they can sustain persistent latent states remains under-explored. The capacity to maintain and manipulate unexpressed, internal representations-analogous to human working memory-is a cornerstone of complex reasoning. In this paper, we formalize and quantify the "Latent State Persistence" (LSP) gap through three novel experiments. First, we utilize a Number Guessing Game, demonstrating that across independent queries, LLMs fail to allocate probability mass to a singular hidden choice, violating a fundamental probabilistic principle. Second, we employ a Yes-No Game to show that as the number of questions increases, LLMs suffer from "concept drift," leading to inevitable self-contradictions due to the lack of LSP. Finally, inspired by Mathematical Mentalism, we task models with tracking transformations on hidden variables, revealing a failure in variable binding and state evolution when the initial state is not explicitly present in the context. Collectively, these findings suggest that LLMs function as reactive post-hoc solvers rather than proactive planners with LSP. Our work provides a framework for evaluating the fidelity of internal representations and highlights a fundamental architectural divergence between autoregressive transformers and human-like cognition.

cs.CL↗

arXiv2026

MMedExpert-R1: Strengthening Multimodal Medical Reasoning via Domain-Specific Adaptation and Clinical Guideline Reinforcement

Medical Vision-Language Models (MedVLMs) excel at perception tasks but struggle with complex clinical reasoning required in real-world scenarios. While reinforcement learning (RL) has been explored to enhance reasoning capabilities, existing approaches face critical mismatches: the scarcity of deep reasoning data, cold-start limits multi-specialty alignment, and standard RL algorithms fail to model clinical reasoning diversity. We propose MMedExpert-R1, a novel reasoning MedVLM that addresses these challenges through domain-specific adaptation and clinical guideline reinforcement. We construct MMedExpert, a high-quality dataset of 10K samples across four specialties with step-by-step reasoning traces. Our Domain-Specific Adaptation (DSA) creates specialty-specific LoRA modules to provide diverse initialization, while Guideline-Based Advantages (GBA) explicitly models different clinical reasoning perspectives to align with real-world diagnostic strategies. Conflict-Aware Capability Integration then merges these specialized experts into a unified agent, ensuring robust multi-specialty alignment. Comprehensive experiments demonstrate state-of-the-art performance, with our 7B model achieving 27.50 on MedXpert-MM and 83.03 on OmniMedVQA, establishing a robust foundation for reliable multimodal medical reasoning systems.

cs.CV↗

arXiv2026

MedEinst: Benchmarking the Einstellung Effect in Medical LLMs through Counterfactual Differential Diagnosis

Despite achieving high accuracy on medical benchmarks, LLMs exhibit the Einstellung Effect in clinical diagnosis--relying on statistical shortcuts rather than patient-specific evidence, causing misdiagnosis in atypical cases. Existing benchmarks fail to detect this critical failure mode. We introduce MedEinst, a counterfactual benchmark with 5,383 paired clinical cases across 49 diseases. Each pair contains a control case and a "trap" case with altered discriminative evidence that flips the diagnosis. We measure susceptibility via Bias Trap Rate--probability of misdiagnosing traps despite correctly diagnosing controls. Extensive Evaluation of 17 LLMs shows frontier models achieve high baseline accuracy but severe bias trap rates. Thus, we propose ECR-Agent, aligning LLM reasoning with Evidence-Based Medicine standard via two components: (1) Dynamic Causal Inference (DCI) performs structured reasoning through dual-pathway perception, dynamic causal graph reasoning across three levels (association, intervention, counterfactual), and evidence audit for final diagnosis; (2) Critic-Driven Graph and Memory Evolution (CGME) iteratively refines the system by storing validated reasoning paths in an exemplar base and consolidating disease-specific knowledge into evolving illness graphs. Source code is to be released.

cs.CL↗

arXiv2026

POLYCHARTQA: Benchmarking Large Vision-Language Models with Multilingual Chart Question Answering

Charts are a universally adopted medium for data communication, yet existing chart understanding benchmarks are overwhelmingly English-centric, limiting their accessibility and relevance to global audiences. To address this limitation, we introduce PolyChartQA, the first large-scale multilingual benchmark for chart question answering, comprising 22,606 charts and 26,151 QA pairs across 10 diverse languages. PolyChartQA is constructed through a scalable pipeline that enables efficient multilingual chart generation via data translation and code reuse, supported by LLM-based translation and rigorous quality control. We systematically evaluate multilingual chart understanding with PolyChartQA on state-of-the-art LVLMs and reveal a significant performance gap between English and other languages, particularly low-resource ones. Additionally, we introduce a companion multilingual chart question answering training set, PolyChartQA-Train, on which fine-tuning LVLMs yields substantial gains in multilingual chart understanding across diverse model sizes and architectures. Together, our benchmark provides a foundation for developing globally inclusive vision-language models capable of understanding charts across diverse linguistic contexts.

cs.CL↗

arXiv2026

Programmable Quantum Anomalous Hall Insulator in Twisted Crystalline Flatbands

The isospin flavors in condensed matters can be continuously broken, forming various symmetry-broken quantum states. In moiré crystals, the competition between different isospin configurations can be effectively tuned by the twist angles and staciking orders. Here we report twisted double rhombohedral-trilayer-gaphene as a new twisted crystalline flatbands system showing rich moiré dependent topological phenomena. In devices with small twist angles, programmable Chern insulators with Chern number C = 3 at integer moiré filling v = 1 have been observed. We have further revealed an exotic hidden order which can quench the Chern insulator as well as multiple first-order transitions between different symmetry-broken phases. Interestly, in the device with a slightly larger twist angle, multiple Chern insulators with C = 1 at fractional moiré fillings including v = 1/4, 1/3 and 1/2 have been observed, whereas the Chern insulator at v = 1 is abscent. Our study demonstrated the twisted flatbands form rhombohedral-multilayer-graphene as a new platform to study tunable high Chern insulators as well as new devices for quantum storage and computation.

cond-mat.mes-hall↗

arXiv2025

AI Deception: Risks, Dynamics, and Controls

As intelligence increases, so does its shadow. AI deception, in which systems induce false beliefs to secure self-beneficial outcomes, has evolved from a speculative concern to an empirically demonstrated risk across language models, AI agents, and emerging frontier systems. This project provides a comprehensive and up-to-date overview of the AI deception field, covering its core concepts, methodologies, genesis, and potential mitigations. First, we identify a formal definition of AI deception, grounded in signaling theory from studies of animal deception. We then review existing empirical studies and associated risks, highlighting deception as a sociotechnical safety challenge. We organize the landscape of AI deception research as a deception cycle, consisting of two key components: deception emergence and deception treatment. Deception emergence reveals the mechanisms underlying AI deception: systems with sufficient capability and incentive potential inevitably engage in deceptive behaviors when triggered by external conditions. Deception treatment, in turn, focuses on detecting and addressing such behaviors. On deception emergence, we analyze incentive foundations across three hierarchical levels and identify three essential capability preconditions required for deception. We further examine contextual triggers, including supervision gaps, distributional shifts, and environmental pressures. On deception treatment, we conclude detection methods covering benchmarks and evaluation protocols in static and interactive settings. Building on the three core factors of deception emergence, we outline potential mitigation strategies and propose auditing approaches that integrate technical, community, and governance efforts to address sociotechnical challenges and future AI risks. To support ongoing work in this area, we release a living resource at www.deceptionsurvey.com.

cs.AI↗