Yu Wang

The paper introduces Kumushi, a root-cause-driven patching agent that significantly improves automated vulnerability repair by focusing LLMs on the true source of bugs, outperforming existing methods and matching commercial agents.

This paper introduces an active traffic analysis method (NATA) and a deep learning framework (BM-Net) to demonstrate that bandwidth perturbations can be used by an adversary to correlate and de-anonymize Tor traffic flows.

The paper introduces TurnGate, a response-aware defense mechanism that detects the earliest turn in a multi-turn dialogue where the accumulated interaction enables a harmful action, significantly improving malicious intent detection.

This paper introduces Agentic Workflow Injection (AWI), a new class of vulnerability in LLM-powered GitHub Actions, and presents TaintAWI, a novel taint-analysis tool that identifies hundreds of exploitable zero-day vulnerabilities.

The paper introduces the Prompt Overflow Attack, demonstrating that guardrail models inspecting truncated or segmented inputs fail to detect malicious instructions that are only actionable when the full, overlong context is provided to the downstream LLM.

The paper analyzes how agentic AI coding assistants can be compromised via prompt injection attacks embedded in external artifacts, turning them into unauthorized execution shells for attackers.

The paper introduces SafeRx-Agent, a knowledge-grounded multi-agent framework that improves medication recommendation accuracy and safety by incorporating fine-grained ATC codes and rigorous safety verification.

VFEAgent is a novel multi-agent framework that automates the entire Finite Element Analysis (FEA) workflow, achieving high success rates in generating complete and physically valid simulations directly from multimodal inputs.

The paper argues that current search agents often verify existing knowledge rather than genuinely searching, and introduces LiveBrowseComp, a new benchmark to measure true evidence-driven discovery.

The paper introduces MOV-Bench, a challenging benchmark for multi-hop audio-visual reasoning, and proposes AOP-Agent, an agentic framework that significantly improves open-source Omni-LLMs' ability to perform active cross-modal perception.

The paper introduces SuiChat-CN, a novel Chinese group-chat benchmark for contextual suicide risk assessment, demonstrating that multi-party conversational context is crucial for accurate detection.

The paper introduces SCALR, a novel framework that generates synthetic user-item interaction data from a source domain to augment a target recommendation domain, significantly improving system performance in A/B tests.

The paper introduces BilliardPhys-Bench, a new benchmark that demonstrates that current multimodal LLMs struggle with complex physical reasoning and predicting object dynamics in simulated environments.

UniScale proposes a unified framework that jointly optimizes model routing and test-time scaling to achieve a superior, fine-grained quality-cost trade-off for large language model inference.

TAPS introduces a target-aware prefix selection method that optimizes the trade-off between draft tree acceptance and verification cost, achieving significant speedups in speculative decoding.

The paper proposes LaSR, a context-aware training paradigm that uses latent reasoning to significantly improve speech recognition, especially for specialized terminology, without adding latency.

UniD$^3$ is a novel Knowledge Graph-enhanced RAG framework that processes vast biomedical literature to systematically extract, organize, and validate comprehensive drug-disease knowledge, achieving high accuracy in structured data generation.

The paper introduces InsightVQA, a large-scale benchmark dataset designed for hierarchical visual question answering that assesses complex emotion understanding and cognitive reasoning beyond simple emotion recognition.

The paper introduces RoleCDE, a novel benchmark that evaluates role-playing agents' ability to resolve conflicts between role-specific values and general alignment constraints, revealing a 'Role Value Decoupling' phenomenon.

The paper introduces RoboTrustBench, a comprehensive benchmark that evaluates the trustworthiness of video world models for robotic manipulation across challenging scenarios, finding that current models fail in complex reasoning and safety checks.