Publications

2 results for Yingbin Liang

DYNAMIC LOSS-BASED SAMPLE REWEIGHTING FOR IMPROVED LARGE LANGUAGE MODEL PRETRAINING
- - Daouda A. Sow
  - Herbert Woisetschläger
  - et al.
- 2025
- ICLR 2025
Conference paper
Random Pruning Over-parameterized Neural Networks Can Improve Generalization: A Training Dynamics Analysis
- - Hongru Yang
  - Yingbin Liang
  - et al.
- 2025
- JMLR
Paper