Zeyu Shi

2 papers · 2025–2026 · 2 conferences · across top CS/AI conferences

Achievements

🌉 Interdisciplinary Bridge 🐝 Cross-Pollinator (13) 🌍 Conference Polyglot (2) 🧭 Keyword Pioneer ❓ The Questioner

Conferences

AAAI (1) ACL (1)

Top co-authors

Haoyi Zhou (2) Jianxin Li (2) Ziming Wang (2) Shiqi Gao (2) Qingyun Sun (2) Tianyu Chen (1)

Keywords

large language model (2) prior knowledge (1) parameter-efficient fine-tuning (1) knowledge boundary (1) calibration error (1) trustworthy ai (1) expected calibration error (1) confidence calibration (1) hessian-guided compensation (1)

Papers

Fine-Tuned LLMs Know They Don’t Know: A Parameter-Efficient Approach to Recovering Honesty AAAI 2026

Towards Objective Fine-tuning: How LLMs’ Prior Knowledge Causes Potential Poor Calibration? ACL 2025