M.S

Master of Science

University of California San Diego

Graduate Research Assistant

Rose Spatiotemporal ML (STL) Lab, UC San Diego

Graduate research in LLM, uncertainty quantification and RL.

TBD

Graduate Student Researcher (GSR)

Dept. of Computer Science and Engineering, UCSD

Summer research support in RoseSTL Lab.

Calibrating LLMs with Semantic-level Reward

arXiv preprint arXiv:2605.15588

Propose Calibration with Semantic Reward (CSR), a semantic-space RL framework that combines correctness reward with semantic calibration reward to improve LLM calibration without relying on verbalized confidence.