Guang Wang

The paper introduces a contextual security framework for LLM agents, defining security properties and reformulating various attacks and defenses based on the context of execution.

This paper introduces a novel framework, the Reasoning Safety Monitor, to detect and prevent logical inconsistencies and adversarial manipulations within the internal reasoning steps of large language models, establishing reasoning safety as a critical security dimension.

The paper proposes M extsuperscript{3}Att, a knowledge-poisoning framework that injects covert misinformation into medical multimodal RAG systems using paired visual data triggers, demonstrating attacks that generate clinically plausible but incorrect diagnoses.

EnergyMamba proposes an uncertainty-aware, graph-enhanced selective state space model to significantly improve both the accuracy and reliability of energy consumption prediction by explicitly modeling spatial dependencies.

The paper introduces MindClaw, a closed-loop framework that enables embodied agents to perform real-time mental-state reasoning and intervene with precision, significantly outperforming standard VLM baselines.

E4GEN introduces an explainable diffusion framework that significantly improves time-series generation by specifically focusing on and controlling the fidelity of extreme events.

The paper introduces CyberGym-E2E, a large-scale, end-to-end benchmark designed to comprehensively evaluate AI agents' capabilities across the entire lifecycle of real-world software vulnerability discovery, proof-of-concept generation, and patch creation.

This paper reveals a denial-of-service vulnerability in LLM-based guardrails for autonomous agents and proposes two attack frameworks.

The paper introduces Epi2Diff, a framework that maps Large Reasoning Models reasoning traces into cognitively grounded episode sequences for human item difficulty prediction, outperforming strong baselines.