arXiv subjects

Hua Wei

Publications and source records attributed to Hua Wei.

At least 19 recordsLinked to original sources

arXiv2026

CHI-Bench: Can AI Agents Automate End-to-End, Long-Horizon, Policy-Rich Healthcare Workflows?

End-to-end automation of realistic healthcare operations stresses three capabilities underrepresented in current benchmarks: policy density, decisions must be grounded in a large library of medical, insurance, and operational rules; Multi-role composition: a single task requires the agent to play multiple roles with handoffs; and multilateral interaction: intermediate workflow steps are multi-turn dialogs, such as peer-to-peer review and patient outreach. We introduce $χ$-Bench, a benchmark of long-horizon healthcare workflows across three domains: provider prior authorization, payer utilization management, and care management. Each task hands the agent a clinical case in a high-fidelity simulator of 20 healthcare apps exposed via 87 MCP tools, which it must drive to a terminal status through tool calls and writing the role's artifacts, guided by a 1,290+ document managed-care operations handbook skill. Across 30 agent harness/models configurations, the best agent resolves only 28.0% of tasks, no agent clears 20% on strict pass^3, and executing all tasks in a single session slumps the performance to 3.8%. These results raise the hypothesis that similar gaps are likely to surface in other policy-dense, role-composed, irreversible enterprise domains.

cs.CL↗

arXiv2026

Steering Vectors as a Training Signal in LLM Post-Training

Steering vectors (SVs) are residual-stream directions that shift the behavior of LLMs when added to hidden states at inference time. While prior work has mainly used SVs as post-hoc controls, we ask whether the same directions can also guide parameter updates during post-training. We study this question across supervised learning, including SFT and distillation, and reinforcement learning with verifiable rewards (RLVR). Our framework considers two intervention regimes: a hidden-state hook during teacher-forced supervised training, and an advantage-repair mechanism on wrong rollouts in GRPO. For each regime, we analyze three design axes, which are the steering direction, its placement, and the operation used to combine it with training. In supervised settings, SV injection improves over plain SFT and distillation, with gains scaling with the off-policy gap and reaching up to a $12\%$ improvement when distilling from thinking-mode teachers. These gains hold across three random seeds. In RL settings, applying a correctness-contrast direction as advantage repair on wrong rollouts, including otherwise zero-gradient all-wrong groups, recovers a learning signal that transfers to held-out math benchmarks and acts as a composable auxiliary signal that adds to existing zero-variance remedies such as entropy-aware advantage shaping. We further identify three failure boundaries. In detail, code generation can break the direction, cross-family layer transfer can break the placement, and inference-time application on an SV-trained checkpoint can break the intervention regime. Together, these results establish steering vectors as a conditional but effective training signal for LLM post-training. Our code is available here.

cs.LG↗

arXiv2026

Detect Before You Disambiguate: Perturbation-Based Input Ambiguity Detection in LLMs

Large language models (LLMs) often answer ambiguous inputs as if they had a single intended meaning. This is problematic because input ambiguity is an upstream source of uncertainty: when a prompt admits multiple valid interpretations, the key question is whether the model should answer directly at all. Existing input uncertainty methods often rely on LLM-generated statements or clarifications, but such transformations may already disambiguate the input before the ambiguity is measured. We propose a label-preserving perturbation framework for input ambiguity detection in which perturbations remain close to the original input while preserving its label. We instantiate this principle with two detectors: a black-box rule-based perturbation library and a white-box importance-guided method using model-internal signals. Experiments across four datasets in open-domain, conversational, and visual QA show that our methods outperform output-side uncertainty baselines and LLM-paraphrase-based perturbation baselines. Our results highlight the importance of detecting ambiguity before attempting to disambiguate it. Our code is available \href{https://github.com/Xiao0o0o/input-ambiguity-detection}{\textcolor{purple}{here}}.

cs.CL↗

arXiv2026

Steering the Teacher: Improving On-Policy Self-Distillation for Thinking Models

On-policy self-distillation (OPSD) trains a student on its own reasoning trajectories using token-level supervision from a reference-conditioned teacher initialized from the same model. However, access to the reference can induce teacher behaviors that undermine independent reasoning in thinking models, leading to reference hallucinations, where the student cites an unavailable reference to justify reasoning steps, and reconsideration suppression, where the teacher discourages verification and revision at uncertain steps. We introduce Steered-Teacher OPSD (ST-OPSD) to jointly address these failure modes through activation steering. ST-OPSD extracts two directions from model-generated reasoning trajectories and applies them to the teacher's hidden states during distillation target computation, suppressing reference-invoking behavior and encouraging reconsideration with marker-family calibration. The student learns from the resulting teacher distributions without being steered. Experiments on AIME and HMMT show that ST-OPSD improves accuracy over vanilla OPSD and other interventions while restoring reconsideration and reference-hallucination rates to base-model levels.

cs.LG↗

arXiv2026

MAVE: A Multi-Agent Framework for Grounded Pedagogical Feedback from Online Class Videos

Timely pedagogical feedback is difficult to scale in online classes, where long and noisy videos make student responses hard to locate and automated systems often produce generic advice weakly grounded in actual classroom behavior. The key challenge is to connect the instructional goal of a specific lesson segment with temporally aligned multimodal evidence of student responses, and then translate the observed gap into an actionable teaching move. We introduce MAVE (Multi-Agents for Video-based Educational Feedback), a multi-agent framework for grounded pedagogical feedback from online-class videos. MAVE decomposes feedback generation into four roles: the Curator structures the lesson and infers expected student activity; the Observer extracts aligned speech, visual, and interaction evidence; the Assessor estimates the activity supported by student behavior; and the Advisor generates gap-conditioned teaching recommendations. To evaluate this capability, we introduce MAVEBench, a benchmark designed to measure whether feedback systems can ground diagnoses in multimodal student evidence, and generate actionable pedagogical feedback. MAVEBench contains 173 knowledge-point segments from approximately 20 hours of authentic online-class videos across STEM and Humanities, together with comprehensive simulation. Experiments show that MAVE outperforms direct prompting, Reflexion, and TextGrad in diagnostic correctness, instructional alignment, and actionability. Our code is available at https://github.com/DaRL-GenAI/MAVE

cs.CY↗

arXiv2026

Diverse Reasoning Paths Matter: Symbolic Equivalence Augmentation for Enhancing RL Exploration

Reinforcement learning with verifiable rewards (RLVR) has emerged as a powerful paradigm for improving the reasoning capabilities of LLMs. However, the role of supervised fine-tuning (SFT) data in enabling effective downstream RL remains underexplored. In this paper, we investigate a key hypothesis: the diversity of reasoning steps in chain-of-thought (CoT) during SFT is a critical enabler of RLVR performance. To test this hypothesis, we introduce EquivAug, a data augmentation framework that uses symbolic equivalence transformations to generate diverse yet provably correct reasoning paths. By converting CoT solutions into formal symbolic representations, applying equivalence transformations, and reconstructing natural-language explanations, EquivAug produces training data that differs in reasoning steps rather than surface-level lexical variation. Experiments on GSM8K and FOLIO across model scales demonstrate that EquivAug consistently achieves the highest reasoning path diversity after SFT and the strongest accuracy after GRPO. Our results establish that CoT diversity during SFT is positively correlated with downstream RLVR performance.

cs.SC↗

arXiv2026

Teaching MLP More Graph Information: A Three-stage Multitask Knowledge Distillation Framework

We study the challenging problem for inference tasks on large-scale graph datasets of Graph Neural Networks: huge time and memory consumption, and try to overcome it by reducing reliance on graph structure. Even though distilling graph knowledge to student MLP is an excellent idea, it faces two major problems of positional information loss and low generalization. To solve the problems, we propose a new three-stage multitask distillation framework. In detail, we use Positional Encoding to capture positional information. Also, we introduce Neural Heat Kernels responsible for graph data processing in GNN and utilize hidden layer outputs matching for better performance of student MLP's hidden layers. To the best of our knowledge, it is the first work to include hidden layer distillation for student MLP on graphs and to combine graph Positional Encoding with MLP. We test its performance and robustness with several settings and draw the conclusion that our work can outperform well with good stability. Code implementation is in https://github.com/lijunxian111/KMP.

cs.LG↗

arXiv2026

When Simulation Lies: A Sim-to-Real Benchmark and Domain-Randomized RL Recipe for Tool-Use Agents

Tool-use language agents are evaluated on benchmarks that assume clean inputs, unambiguous tool registries, and reliable APIs. Real deployments violate all these assumptions: user typos propagate into hallucinated tool names, a misconfigured request timeout can stall an agent indefinitely, and duplicate tool names across servers can freeze an SDK. We study these failures as a sim-to-real gap in the tool-use partially observable Markov decision process (POMDP), where deployment noise enters through the observation, action space, reward-relevant metadata, or transition dynamics. We introduce RobustBench-TC, a benchmark with 22 perturbation types organized by these four POMDP components, each grounded in a verified GitHub issue or documented tool-calling failure. Across 21 models from 1.5B to 32B parameters (including the closed-source o4-mini), the robustness profile is sharply uneven: observation perturbations reduce accuracy by less than 5%, while reward-relevant and transition perturbations reduce accuracy by roughly 40% and 30%, respectively; scale alone does not close these gaps. We then propose ToolRL-DR, a domain-randomization reinforcement learning (RL) recipe that trains a tool-use agent on perturbation-augmented trajectories spanning the three statically encodable POMDP components. On a 3B backbone, ToolRL-DR-Full retains roughly three-quarters of clean accuracy and reaches an aggregate perturbed accuracy comparable to open-source 14B function-calling baselines while substantially narrowing the gap to o4-mini. It closes approximately 27% of the Transition gap despite never seeing transition perturbations in training, suggesting that RL on adversarial static tool-use inputs induces a more persistent retry policy that transfers to unseen runtime failures. The dataset, code and benchmark leaderboard are publicly available.

cs.AI↗

arXiv2026

GeoReform: Reflective Formalization Evolution for Multimodal Geometry Problem Solving

Multimodal large language models (MLLMs) often struggle to identify and use geometric relations in diagrams. Recent methods address this challenge by converting geometric entities, relations, and constraints into explicit textual representations for the model to reason over. However, effective formalization is highly non-trivial: on Geometry3K, structure injection fixes 28 errors but introduces 13 new ones among 200 examples. Redundant relations can distract the model, while ambiguous references to diagram elements can lead it to apply constraints incorrectly. This suggests that the key challenge is not merely extracting more geometric facts, but organizing them into representations that support downstream reasoning. To fully exploit the power of formalization, we further propose GeoReform, a reflective formalization evolution framework that treats formalization as an optimizable policy rather than a fixed parser output. GeoReform executes the full reasoning pipeline, collects failed rollouts, diagnoses defects in the current representation, and mutates the policy to better select, ground, group, and present geometric entities, relations, constraints, and targets. On Geometry3K, GeoReform improves Qwen3VL-2B accuracy from 42.0\% to 56.0\%. Extensive experiments and analyses across geometry reasoning benchmarks demonstrate that effective formalization is crucial for improving multimodal geometry reasoning.

cs.AI↗

arXiv2026

Generative AI for Autonomous Driving: Frontiers and Opportunities

Generative Artificial Intelligence (GenAI) constitutes a transformative technological wave that reconfigures industries through its unparalleled capabilities for content creation, reasoning, planning, and multimodal understanding. This revolutionary force offers the most promising path yet toward solving one of engineering's grandest challenges: achieving reliable, fully autonomous driving, particularly the pursuit of Level 5 autonomy. This survey delivers a comprehensive and critical synthesis of the emerging role of GenAI across the autonomous driving stack. We begin by distilling the principles and trade-offs of modern generative modeling, encompassing VAEs, GANs, Diffusion Models, and Large Language Models (LLMs). We then map their frontier applications in image, LiDAR, trajectory, occupancy, video generation as well as LLM-guided reasoning and decision making. We categorize practical applications, such as synthetic data workflows, end-to-end driving strategies, high-fidelity digital twin systems, smart transportation networks, and cross-domain transfer to embodied AI. We identify key obstacles and possibilities such as comprehensive generalization across rare cases, evaluation and safety checks, budget-limited implementation, regulatory compliance, ethical concerns, and environmental effects, while proposing research plans across theoretical assurances, trust metrics, transport integration, and socio-technical influence. By unifying these threads, the survey provides a forward-looking reference for researchers, engineers, and policymakers navigating the convergence of generative AI and advanced autonomous mobility. An actively maintained repository of cited works is available at https://github.com/taco-group/GenAI4AD.

cs.CV↗

arXiv2026

COLLATOR: Compositional Multi-Agent Orchestration with Counterfactual Reinforcement Learning

Large language models (LLMs) provide a flexible foundation for multi-agent systems, but their effectiveness and computational cost depend critically on orchestration design. Across different tasks, role design, capacity assignment, and dependency construction jointly affect both solution quality and execution efficiency. Existing approaches automate parts of this design process, yet they often optimize these decisions partially or sequentially, and rely on execution-level feedback that provides limited credit assignment for local orchestration decisions. We propose LEMON (Learning Executable Multi-Agent Orchestration via Counterfactual Reinforcement Learning), an LLM-based orchestrator that learns to design efficient multi-agent orchestration. Given a task, LEMON designs a unified orchestration specification that composes customized agent duties, capacity levels, and dependency relations. To train the orchestrator, we augment the orchestration-level Group Relative Policy Optimization (GRPO) objective with a localized counterfactual credit signal that edits role, capacity, or dependency fields and applies the resulting reward contrast only to the edited spans. Experiments on six reasoning and coding benchmarks, including MMLU, GSM8K, AQuA, MultiArith, SVAMP, and HumanEval, show that LEMON achieves the best average performance among evaluated multi-agent orchestration methods while improving the accuracy-token trade-off.

cs.AI↗

arXiv2026

Auditable Agents

LLM agents call tools, query databases, delegate tasks, and trigger external side effects. Once an agent system can act in the world, the question is no longer only whether harmful actions can be prevented--it is whether those actions remain answerable after deployment. We distinguish accountability (the ability to determine compliance and assign responsibility), auditability (the system property that makes accountability possible), and auditing (the process of reconstructing behavior from trustworthy evidence). Our claim is direct: no agent system can be accountable without auditability. To make this operational, we define five dimensions of agent auditability, i.e., action recoverability, lifecycle coverage, policy checkability, responsibility attribution, and evidence integrity, and identify three mechanism classes (detect, enforce, recover) whose temporal information-and-intervention constraints explain why, in practice, no single approach suffices. We support the position with layered evidence rather than a single benchmark: lower-bound ecosystem measurements suggest that even basic security prerequisites for auditability are widely unmet (617 security findings across six prominent open-source projects); runtime feasibility results show that pre-execution mediation with tamper-evident records adds only 8.3 ms median overhead; and controlled recovery experiments show that responsibility-relevant information can be partially recovered even when conventional logs are missing. We propose an Auditability Card for agent systems and identify six open research problems organized by mechanism class.

cs.AI↗

arXiv2026

PlantShade: Predicting Plant Shadows for Lighting-Aware Robotic Agricultural Operation

Plant growth and agricultural production form the foundation of a country's sustainable development and directly impact human livelihoods. Recent advances in frontier artificial intelligence have enabled scientific agriculture with strong potential to improve crop productivity. In this paper, we identify the importance and inherent complexity of plant shade simulation, as shading is a critical factor influencing plant growth. To advance this field and promote broader societal benefits, we focus on two main contributions. First, we introduce a comprehensive plant growth and shade dataset covering four plant species, including soybean, tomato, sugarbeet, and strawberry. The dataset includes top-down viewpoints with a supplementary light along a circular trajectory, casting dynamic shadows across multiple growth stages and diverse observation complexities. Second, we propose generative shade simulation based on diffusion models, enabling realistic shade generation for unseen plants and supporting downstream robotic tasks such as perception, lighting control, and view planning. The model incorporates temporal conditioning to facilitate flexible shade simulation across different time stages. We conduct both quantitative and qualitative evaluations to assess model performance. This work provides a foundational study for plant-aware shade modeling and has meaningful implications for broader agricultural and robotic applications.

cs.RO↗

arXiv2026

From Rubrics to Reliable Scores: Evidence-Grounded Text Evaluation with LLM Judges

Rubric-based text evaluation increasingly relies on large language models (LLMs) as scalable judges, yet frozen black-box models can interpret the same criteria inconsistently, produce score attributions that are difficult to audit, and map judgments poorly onto human scoring scales. We define this challenge as criteria transfer: translating human rubric intent into a stable, auditable inference-time scoring protocol. We introduce Rulers, which locks a task-level rubric specification, executes it through structured, evidence-grounded judgments, and calibrates the resulting signals to human score boundaries. Across four rubric-governed benchmarks and multiple frozen backbone models, Rulers achieves stronger agreement with human scores in most evaluated settings, while better matching empirical score distributions and remaining more stable under semantically equivalent rubric perturbations. Calibration controls and component ablations show that these gains cannot be attributed to post-hoc alignment alone, but depend on the combination of fixed criteria, traceable evidence, and calibrated score interpretation. These findings suggest that reliable LLM judging requires faithfully operationalizing human evaluation standards rather than relying on prompt-level scoring alone. Our code is available at https://github.com/LabRAI/Rulers.git.

cs.CL↗

arXiv2026

Sim2Signal: Sim-to-Real Benchmarks for Traffic Signal Control

Reinforcement learning achieves strong traffic signal control performance in simulation, yet policies trained in simulators often fail once deployed in the real world, a failure known as the Sim-to-Real gap. When RL is applied to traffic signal control, this gap arises from several sources: sensing, action execution, traffic dynamics, and the control objective. Their relative impact and the reliability of existing Sim-to-Real mitigation methods remain insufficiently understood, and the field lacks a standard benchmark for systematically measuring the gap and evaluating mitigation methods. We present Sim2Signal, a benchmark that decomposes the Sim-to-Real gap into observation, action, transition, and reward gaps, corresponding to mismatches in the four components of the underlying MDP, and induces each gap in isolation under a shared protocol. We evaluate 18 mitigation methods on 2 base controllers, across 33 gap settings and 10 calibrated networks built from 5 real-world locations. We find that direct transfer consistently degrades performance across all four gap sources, but the severity of the degradation does not predict the effectiveness of mitigation. Instead, mitigation effectiveness depends strongly on the network and gap setting: outside the action gap, a method that helps in one case may fail in another. The most effective methods generally estimate what the gap changes, rather than make the policy insensitive through domain randomization or invariant representations. Our code is available at https://github.com/DaRL-LibSignal/Sim2Signal

cs.LG↗

arXiv2026

MASkills: Continual Skills Optimization for Multi-Agent LLM Systems

LLM-based multi-agent systems have shown strong performance on complex tasks, yet continual improvement from interaction experience remains challenging. Existing self-reflection methods build experience memories, but memories are mostly hard to invoke, refine, or scale, while agent skills offer a more actionable unit: structured procedural knowledge that specifies when to act, how to act, and which resources or tools to use. We introduce MASkills, a continual learning framework that optimizes multi-agent LLM systems through agent skills. MASkills presents a new agent-optimization pipeline that integrates skill-conditioned credit assignment, hierarchical credit aggregation, and momentum-smoothed optimization, enabling agent skill libraries to evolve through refinement, induction, consolidation, and pruning. Experiments on HotpotQA, LoCoMo, and GAIA demonstrate the effectiveness of MASkills across multiple agentic tasks. Our code is available at https://github.com/DaRL-GenAI/MASkills

cs.AI↗

arXiv2026

AndroidReality: How Far Are Mobile Agents from the Real World?

Mobile agents have achieved promising results on clean online benchmarks such as AndroidWorld, yet their performance often degrades sharply in real-world deployment due to environmental variations and imperfect interface conditions. In this work, we introduce AndroidReality, a perturbation-based framework for evaluating and improving the robustness of mobile agents. Through a Markov Decision Process (MDP) perspective, we organize real-world interface variability into a principled taxonomy of perturbations along three axes: state, transition, and action. Guided by this taxonomy, we build a perturbed mobile benchmark on top of AndroidWorld with realistic and controllable perturbation injections, enabling systematic robustness evaluation of mobile agents. Our evaluation reveals substantial robustness gaps and four recurring error categories, motivating a simple training-free Test-Time Introspective Recovery (TTIR) mechanism that mitigates these failures on both perturbed and clean settings. Together, these results position robustness as a missing dimension in mobile agent evaluation and establish benchmark perturbation as an effective tool for both stress testing and surfacing latent weaknesses of mobile agents.

cs.AI↗

arXiv2026

Agentic Commerce World: An Auditable and Verifiable Environment for Vibe Commerce

In vibe coding, people describe software in natural language and delegate implementation to AI agents. By analogy, vibe commerce allows people to express buying or selling goals in natural language and delegate the corresponding tasks to agents. Commerce, however, requires independently controlled Buyer and Merchant agents to interact in a shared market while preserving their private objectives and distinct authority. We introduce Agentic Commerce World (ACWorld), an environment for evaluating such agents across ongoing transactions. Through its Vibe Commerce Protocol (VCP), ACWorld validates agent actions before updating shared transaction state and records the resulting interactions, making agent behavior auditable and evaluation reproducible. The ACWorld Benchmark contains a 200-task capability-coverage track and a 60-task large-catalog track that searches 785,022 transactable listings. Across ten models, mean scores range from 65.9% to 85.6% and from 56.1% to 91.4%, respectively. Our analysis shows that process-level evidence is necessary: final state alone can miss evaluated errors, incomplete trajectories still retain useful process signals, and large-catalog tasks expose bottlenecks across stages.

cs.AI↗