Sept. 2026: Our paper Counterfactual Debugging the World Model Transfer Gap is accepted at NeurIPS 2026!- May 2026: Two papers on imitation learning and causal credit assignment are accepted at RLC 2026!
- Apr. 2026: Our paper Causal Flow Q-Learning for Robust Offline Reinforcement Learning is accepted at ICML 2026!
- Feb. 2026: We will present our new work Causal Flow Q-Learning for Robust Offline Reinforcement Learning at the Princeton Robotics Inaugural Symposium on April 20th 2026.
- Jan. 2026: I am elected to be a MATS scholar (26 Spring, 7% acceptance rate) to work on AI safety and alignment!
- Jan. 2026: Our paper Confounding Robust Reinforcement Learning: A Causal Approach is accepted at Amazon Trusted AI Symposium 2026!
- Dec. 2025: Our paper Confounding Robust Continuous Control via Automatic Reward Shaping is accepted at AAMAS 2026!
- Oct. 2025: I have been selected as a “Top Reviewer” (8%, 1958/24429) for NeurIPS 2025!
- Sept. 2025: Our paper Confounding Robust Reinforcement Learning: A Causal Approach is accepted at NeurIPS 2025!
- Aug. 2025: I wrapped up my summer internship at Uber as a Research Scientist.
- May 2025: Our paper Automatic Reward Shaping from Confounded Offline Data is accepted at ICML 2025!
- April 2025: I am co-organizing the First Causal Reinforcement Learning Workshop at RLC 2025, please submit your work to CRLW!
Mingxuan Li
ml@cs.columbia.edu
© 2026