Published several preprints such as 'Speculative Actions: A Lossless Framework for Faster Agentic Systems' and 'Exchangeable Sequence Models Can Naturally Quantify Uncertainty Over Latent Concepts'. Also, had work accepted to ICLR 2025. The paper 'Differences-in-Neighbors for Network Interference in Experiments' was a finalist for the Jeff McGill Student Paper Award 2025.
Research Experience
Involved in multiple research projects including Algorithmic Learning for Sequence Models and the development of SynthTools framework.
Education
B.A. in Mathematics from Princeton University; Ph.D. candidate in the Decision, Risk, and Operations department at Columbia University, advised by Prof. Hongseok Namkoong and Prof. Tianyi Peng.
Background
A fourth year Ph.D. student in the Decision, Risk, and Operations department at Columbia University. Interested in training and deploying agentic systems, with a focus on developing AI agents for real-world sequential decision-making under uncertainty.
Miscellany
Reviewer for ICLR 2026, ICLR 2025, AISTATS 2026, and NeurIPS 2024.