From Prompt Injection to Persistent Control: Defending Agentic Harness Against Trojan Backdoors
Paper • 2605.31042 • Published • 19
Xinyu Yang
yxy919
AI & ML interests
None yet
Recent Activity
upvoted a paper about 1 month ago
Toward Generalist Autonomous Research via Hypothesis-Tree Refinement authored a paper about 1 month ago
From Prompt Injection to Persistent Control: Defending Agentic Harness Against Trojan Backdoors upvoted a paper about 1 month ago
Towards Verifiable Multimodal Deep Research: A Multi-Agent Harness for Interleaved Report GenerationOrganizations
None yet
authored a
paper about 1 month ago