Faculty Profile: Prof. Lei Li — CMU LTI
Faculty Profile: Prof. Lei Li — CMU LTI
Position: Associate Professor Institution: Carnegie Mellon University, Language Technologies Institute Website: https://lileicc.github.io/ Report date: 2026-06-12
Research Focus
LLM generation quality, multilingual NLP, machine translation, text watermarking, LLM evaluation and benchmarking, structured text generation.
Academic Profile
- PhD: Carnegie Mellon University
- Previously at ByteDance AI Lab (research director) and Amazon
- Publications in ACL, EMNLP, NeurIPS, ICML
- Organizing/program committee roles in major NLP venues
Key Publications
| Paper | Venue | Focus |
|---|---|---|
| Text watermarking for LLMs | 2023–2024 | IP protection and traceability for generated text |
| Multilingual machine translation | ACL/EMNLP | Cross-lingual transfer |
| Structured text generation | NeurIPS/ICML | Controllable generation |
Fit with Weijia Zhang
| Dimension | Assessment |
|---|---|
| LLM generation / controllability | ✅ Core focus |
| Multilingual NLP | ✅ Strong |
| Text watermarking / attribution | ✅ Novel angle |
| NLP agents | ⚠️ Adjacent |
| GUI / VLM agents | ❌ Not his area |
| RL post-training | ❌ Not primary |
Verdict
P3 套磁。 Relevant if Weijia is interested in LLM generation quality, multilingual capabilities, or watermarking. Less central to agentic AI direction. Industry background (ByteDance, Amazon) provides practical perspective.