Scholar
Quzhe Huang
Google Scholar ID: OGITFL4AAAAJ
Ph.D. Student, Peking University
Citations & Impact
Citations
931
H-index
12
i10-index
14
Publications
20
Co-authors
8
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Publications: 'Let Real Users Decide: Evaluate Role-Play Chatbot with User Simulator' (Under Review), 'Automating Legal Concept Interpretation with LLMs: Retrieval, Generation, and Evaluation' (Under Review), 'JUREX-4E: Juridical Expert-Annotated Four-Element Knowledge Base for Legal Reasoning' (Under Review), 'Unlocking the Potential of Model Merging for Low-Resource Languages' (Findings of EMNLP 2024), 'Harder Tasks Need More Experts: Dynamic Routing in MoE Models' (ACL 2024), 'Lawyer LLaMA: Enhancing LLMs with Legal Knowledge' (Arxiv); Open-sourced code and datasets.
Research Experience
- LLM researcher at Kuaishou; Participated in projects covering the full training procedure of LLMs, including Pretraining, SFT, and RLHF, often in a leadership role; Developed the first Chinese Legal LLM, Lawyer LLaMA; Contributed to the development of a unified language and vision pretraining framework.
Education
- Bachelor's and Ph.D. degrees from Peking University, advised by Prof. Yansong Feng and Prof. Dongyan Zhao.
Background
- Research Interests: Improving LLMs with real human feedback; Specialization: LLM training, domain-specific LLMs, and multimodal LLMs.
Miscellany
- Personal interests not mentioned
