News

2026 I will join UCLA as a Ph.D. student in Computer Science.
2025 Our paper on the implicit bias of per-sample Adam will appear at ICLR 2026.
2025 Our paper on schedule-free methods for language model training will appear at NeurIPS 2025.