12 papers
Practice Makes Unsafe: Skill Misevolution in Self-Improving LLM Agents
Xutao Mao, Liangjie Zhao, Xiang Zheng +1
Self-improving LLM agents convert successful trajectories into persistent cross-task state. An unsafe success can thereby become reusable policy after its triggering input disappea…
Agents Don't Just Agree, They Remember: Benchmarking Persistent Sycophancy in Stateful Personal Agents
Xutao Mao, Liangjie Zhao, Leyao Wang +6
Stateful personal agents increasingly maintain long-term user profiles, episodic memories, and reusable skills. This persistence turns conversational sycophancy into a state-writin…
Agent Hacks Agent: Autoresearch for Production-Agent Red-Teaming
Xutao Mao, Xiang Zheng, Cong Wang
Production LLM agents such as Claude Code and Codex operate over untrusted content, files, commands, and workspace state, making safety failures directly actionable. Red-teaming mu…
Safety in Embodied AI: A Survey of Risks, Attacks, and Defenses
Xiao Li, Xiang Zheng, Yifeng Gao +35
Embodied Artificial Intelligence (Embodied AI) integrates perception, cognition, planning, and interaction into agents that operate in open-world, safety-critical environments. As…
STARE: Step-wise Temporal Alignment and Red-teaming Engine for Multi-modal Toxicity Attack
Xutao Mao, Liangjie Zhao, Tao Liu +3
Red-teaming Vision-Language Models is essential for identifying vulnerabilities where adversarial image-text inputs trigger toxic outputs. Existing approaches treat image generatio…
HazardArena: Evaluating Semantic Safety in Vision-Language-Action Models
Zixing Chen, Yifeng Gao, Li Wang +8
Vision-Language-Action (VLA) models inherit rich world knowledge from vision-language backbones and acquire executable skills via action demonstrations. However, existing evaluatio…