Scholar
Ziniu Li
Google Scholar ID: 80UnKQQAAAAJ
The Chinese University of Hong Kong, Shenzhen
Citations & Impact
Citations
803
H-index
14
i10-index
16
Publications
20
Co-authors
16
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Best-Paper Runner-Up, NeurIPS 2024 FITML Workshop
- Oral presentations at ICLR 2024 Tiny Paper Track, UAI 2023, and NeurIPS 2021 EcoRL Workshop
- Spotlight presentation at NeurIPS 2023
- Research on large language models includes: data selection (NeurIPS 2023 Spotlight), diversity-preserving supervised fine-tuning (ICLR 2025; NeurIPS 2024 FITML Workshop Best Paper Runner-up), generalization of RLHF (ICLR 2024 Tiny Paper Oral), computationally efficient RLHF (ICML 2024), and hallucination mitigation (ICLR 2025)
- In imitation and reinforcement learning: theoretical work on sample complexity (NeurIPS 2020, TPAMI 2021, UAI 2023 Oral), efficient exploration (ICLR 2022, NeurIPS 2021 EcoRL Workshop Oral, DAI 2020), and applications in robotics (ICLR 2024 Blog) and signal processing (TSP 2024)
- Collaborative work on optimization: understanding Adam in Transformers (NeurIPS 2024), memory-efficient optimizers (ICLR 2025), zero-order optimization (IJCAI 2020), and prompt-tuning (EMNLP 2024)
- Served as reviewer for NeurIPS (Top Reviewer), ICML (Outstanding Reviewer), and ICLR (Highlighted Reviewer)
