Scholar
Lei Li
Google Scholar ID: BYXqAlwAAAAJ
Associate Professor, School of Computer Science, Carnegie Mellon University
Citations & Impact
Citations
21,561
H-index
60
i10-index
156
Publications
20
Co-authors
162
list available
Contact
Publications
Resume (English only)
Academic Achievements
- CMU team achieved Top-1 scores in IWSLT 2025 simultaneous speech translation competition (English-Chinese and English-German)
- 3 papers accepted by ACL 2025 (1 Main + 2 Findings) on simultaneous speech translation, LLM watermark detection, and multilingual translation for 400+ languages
- 3 papers accepted by ICML 2025 on protein-protein complex design, LLM copyright, and safety
- 6 papers accepted by NAACL 2025 (4 Main + 2 Findings) on machine translation with LLMs, scaling LLM inference-time compute, and efficient finetuning
- 4 papers accepted by ICLR 2025 on LLM watermarking (permute-flip), LLM reasoning, and speculative LLM knowledge distillation
- 2 papers accepted by NeurIPS 2024 on GenAI watermarking and LLM reasoning in non-English languages
- 7 papers accepted by EMNLP 2024 (3 Main + 3 Findings + 1 system demo) on LLM preference learning, multilingual LLM training (LLaMAX), evaluation, and MT analysis tools
- 1 paper accepted by KDD 2024 on counterfactual molecule property explanation via reinforcement learning
- 3 papers accepted by ACL 2024 (including 2 Findings) on self-bias in LLMs, LLMs for translating any language (including endangered ones), and analyzing LLaMA’s multilingual capability through vocabulary
- 3 papers accepted by ICML 2024 on protein design (SurfPro and EnzyGen) and copyright content detection in LLM training
- 2 papers accepted by NAACL 2024 Findings on LLM refinement and evaluating LLMs for multilingual translation
- 1 paper accepted by ICLR 2024 on LLM watermarking for detecting AI-generated text
- 1 paper accepted by AAAI 2024 on LLMs for wildlife conservation
- 4 papers accepted by EMNLP 2023 (including 2 Findings) on using Study Assistant to correct LLM mistakes, InstructScore, automatic planning for decision-making tasks, and adapting NLP encoders to generators
- 2 papers accepted by NeurIPS 2023 on LLM code generation for algorithmic problems and assessing knowledge in LLMs
- 1 paper accepted by KDD 2023 on designing antibacterial peptides
- 5 papers accepted by ACL 2023 on speech translation, evaluation metric SEScore2, and machine translation
