 Weak-to-Strong Generalization Even in Random Feature Networks, ProvablyACM Transactions on Software Engineering and Methodology (TOSEM), 2024 |
 High-dimensional Analysis of Knowledge Distillation: Weak-to-Strong Generalization and Scaling LawsInternational Conference on Learning Representations (ICLR), 2024 |
 Provable Weak-to-Strong Generalization via Benign OverfittingInternational Conference on Learning Representations (ICLR), 2024 |
 Your Weak LLM is Secretly a Strong Teacher for AlignmentInternational Conference on Learning Representations (ICLR), 2024 |
 A Theory of Non-Linear Feature Learning with One Gradient Step in Two-Layer Neural NetworksInternational Conference on Machine Learning (ICML), 2023 |
