Scholar

Mingxuan Li

Google Scholar ID: V4MEw70AAAAJ

Columbia University

Causal InferenceReinforcement LearningAdversarial Robustness

Citations & Impact

All-time

Citations

H-index

i10-index

Publications

Co-authors

list available

Contact

Publications

2 items

2026

Cited

2026

Cited

Resume (English only)

Academic Achievements

- Publications:
- 'Confounding Robust Reinforcement Learning: A Causal Approach' accepted at NeurIPS 2025
- 'Automatic Reward Shaping from Confounded Offline Data' accepted at ICML 2025
- 'Causally Aligned Curriculum Learning' accepted at ICLR 2024
- 'Learning Generalizable Behavior via Visual Rewrite Rules' presented at AAAI-22 Workshop on Reinforcement Learning in Games
- 'Towards Sample Efficient Agents through Algorithmic Alignment' presented at AAAI-21 Student Abstract and Poster Program
- 'Interpretability is a Kind of Safety: An Interpreter-based Ensemble for Adversary Defense' presented at KDD-20 (Oral)
- Talks:
- April 2025: Talk at NSF Causal Decision Making Seminar
- March 2025: Guest lecture at UCI CompSci 295, Causal Inference for Reinforcement Learning
- Organization:
- Organizer of The Causal Reinforcement Learning Workshop (RLC 2025)
- Reviewer Experience:
- ICML 2025; NeurIPS 2023, 2024, 2025; ICLR 2024, 2025, 2026; AISTATS 2025, 2026; AAAI 2026; Journal of Machine Learning Research (JMLR); International Journal of Robotics Research (IJRR)

Research Experience

Education

Background

- Research Interests: Intersection of causal inference and reinforcement learning, especially in building a causally aligned, generalizable, and sample-efficient agent
- Professional Field: Causal Reinforcement Learning
- Brief Introduction: In the era of large language models, believes that a causally aligned agentic system is more important than ever. Focused on designing practical and scalable causal tools to mitigate problems in language model training/testing.

Miscellany