My research broadly spans topics in machine learning and formal methods. My goal is to ensure reasoning in AI systems is reliably accurate, efficient and adaptive. Specifically, I'm interested in:
- Correctness and efficiency of LLM reasoning;
- Verifiability and safety of model behaviors.
If you believe I can help you in any way or you'd just like to chat, message me at:
nihal.jain@columbia.edu.
I'm also available at the following channels:
