Research focuses on Large Language Models (LLMs) and reasoning
Aims to build intelligent systems capable of complex reasoning across diverse domains
Advances LLM reasoning through novel training paradigms combining reinforcement learning and structured feedback
Develops rigorous benchmarks for evaluating machine intelligence in scientific reasoning
Has foundational work in causal discovery and inference methods for open-world observational data
Research includes causal discovery algorithms, transfer learning, and applications in neuro-behavioral data analysis and medical causal effect inference