Han Luo
I am a second-year undergraduate student in Computer Science at the University of Leeds, UK, and Southwest Jiaotong University, China. My research interests lie in large language model agents, multi-agent systems, and responsible AI, with a particular focus on how language models behave in complex, interactive environments.
Selected publications
View all >
Agentic Abstention
A benchmark and context-engineering study of when LLM agents should stop acting and abstain.
Plan in Sandbox
A physics-grounded sandbox framework for learning abstracted embodied navigation experience.
SPASM
A stability-first framework for persona-driven LLM-LLM dialogue simulation with egocentric context projection.
DialogGuard
A practitioner-facing interface for multi-agent psychosocial safety evaluation of sensitive LLM responses.
News
Jun 27, 2026
Agentic Abstention preprint released
Our paper, Agentic Abstention: Do Agents Know When to Stop Instead of Act?, is now available as arXiv:2606.28733.
May 11, 2026
SAGE arXiv preprint released
Our ICML 2026 paper, Plan in Sandbox, Navigate in Open Worlds, is now available as arXiv:2605.10118.
May 1, 2026
Paper accepted to ICML 2026
Our collaborative paper, Plan in Sandbox, Navigate in Open Worlds, was accepted to ICML 2026.
Apr 27, 2026
Two papers accepted to ACL 2026
SPASM was accepted to Findings of ACL 2026, and DialogGuard was accepted to ACL 2026 System Demonstrations.
Upcoming
May 2026 Conference or talk City
Jun 2026 Research visit Institution