News

Feb. 2024 New paper Structure-agnostic Optimality of Doubly Robust Learning for Treatment Effect Estimation with Vasilis Syrgkanis posted on ArXiv! Fantastic results showing that first-order debiasing is optimal for black-box ML-based nuisance estimators.

Jan. 2024 Our paper Dichotomy of Early and Late Phase Implicit Biases Can Provably Induce Grokking is accepted by ICLR 2024!

Nov. 2023 New paper Dichotomy of Early and Late Phase Implicit Biases Can Provably Induce Grokking with Keifeng Lyu, Zhiyuan Li, Simon S. Du, Jason D. Lee and Wei Hu posted on ArXiv! Welcome any feedback or comments.

Nov. 2023 New paper Learning Causal Representations from General Environments: Identifiability and Intrinsic Ambiguity with Vasilis Syrgkanis posted on ArXiv! Welcome any feedback or comments.

Oct. 2023 Give a talk at the 2023 INFORMS Annual Meeting based on our incremental learning paper. Had a wonderful time in Phoenix!

Sept. 2023 Arriving at Stanford and start my Ph.D. journey!

Jul. 2023 Graduate from Peking University! Will miss the time spent there.

Apr. 2023 Our paper Understanding Incremental Learning of Gradient Descent: A Fine-grained Analysis of Matrix Sensing is accepted by ICML 2023!

Jan. 2023 Our Minimax Optimal Kernel Operator Learning via Multilevel Training is accepted by ICLR 2023 as notable 25% (spotlight presentation)!

Oct. 2022 Our Minimax Optimal Kernel Operator Learning via Multilevel Training is accepted by the NeurIPS 2022 AI4Science workshop.

Sept. 2022 Our new paper on minimax optimal kernel operator learning has been posted on ArXiv.

Sept. 2022 Personal website created!