Scholar
Luxi He
Google Scholar ID: 9xrZTooAAAAJ
Department of Computer Science, Princeton University
Citations & Impact
Citations
400
H-index
8
i10-index
8
Publications
14
Co-authors
0
Publications
Cited
0
Resume (English only)
Academic Achievements
- Paper 'Statutory Construction and Interpretation for Artificial Intelligence' accepted at NeurIPS RegML Workshop (Oral, 2025)
- Paper 'The Model Hears You: Audio Language Model Deployments Should Consider the Principle of Least Privilege' accepted at AIES 2025 (Oral)
- Presented CharXiv at NeurIPS 2024 and gave an oral presentation at the EvalEval Workshop
- Gave a spotlight presentation at ICML 2024 GenLaw Workshop on 'Fantastic Copyrighted Beasts'
- Published work on interpreting and constructing natural language rules for AI, accompanied by X thread, blog post, and policy brief
Background
- Third-year Ph.D. student in Computer Science at Princeton University
- Research focuses on understanding language models and improving their alignment and safety
- Particularly interested in the impact of data throughout the language model lifecycle
- Aims to make language models more reliable and trustworthy
- Recently exploring human-LLM alignment and collaboration from first principles
- Motivated to bridge technology and policy by integrating insights from both domains
Co-authors