| Oct 2026 | Two papers accepted to EMNLP 2026: ADWIN (Main) and ORBIT (Findings). |
| Sep 2026 | πPPO: Privy to the Foil is available on arXiv. |
| May 2026 | ADWIN, adaptive windows for horizon-aware on-policy distillation, is out on arXiv. |
| Jan 2026 | ORBIT, on-policy exploration-exploitation for controllable multi-budget reasoning, is out on arXiv. |
| Nov 2025 | Composable Cross-prompt Essay Scoring by Merging Models appeared at EMNLP 2025 (Main Conference, Oral). |
| Sep 2025 | Started my Ph.D. at Peking University, advised by Prof. Yunfang Wu. |
| Jul 2025 | Joined the RL Team at Tencent Hunyuan as a research intern. |