About Me
Welcome to my homepage! I am a postdoctoral research associate at the Machine Learning Department of Carnegie Mellon University, working with Prof. Aarti Singh and Prof. Bryan Wilder. Before joining CMU, I got my PhD in Computer Science at UC Santa Barbara in 2024, advised by Prof. Yu-Xiang Wang. Even before that, I got my B.S. from Tsinghua University in 2019, advised by Prof. Guoqi Li.
Research Interests: I am broadly interested in online learning under uncertainty:
- Theoretic foundations:
- Generative decision making (see below).
- Dynamic pricing under constraints.
- Online optimization with non-convexity.
- Applications:
- AI-driven healthcare decision support.
- In-context decision planning for finance and markets .
About Generative Decision-Making: With the help of Generative AI, now decision-makers may generate new and customized actions beyond learning on existing action sets. Beyond the traditional balance of exploration and exploitation, my works introduce a third dimension –creation– by developing provable algorithms to actively decide when to generate and what to generate given streams of contexts coming on-the-fly. See our recent ICLR-26 paper here as a preliminary demo.
News and Events
May 2026. Our manuscript “Optimal Contextual Pricing under Agnostic Non-Lipschitz Demand” is posted on ArXiv. This result finally closes the long-existing regret gap of feature-based dynamic pricing with linear & noisy valuation. Thanks to Yu-Xiang!
Mar 2026. Our manyscript is posted on ArXiv. This work comprehensively documents the procedure of our developing a maternal healthcare chatbot with a RAG-LLM framework, along with carefully-designed safety guardrails and calibrated evaluation criterias. Thanks to Smriti and Vidhi for the consistent efforts on this pipeline development, and to all collaborators from the U.S. and India.
Jan 2026. Our paper “Online Decision Making with Generative Action Sets” is accepted by ICLR 2026! Thanks to my collaborators Vidhi, and my advisors Bryan and Aarti!
Oct 2025. Glad to receive NeurIPS Top Reviewer for the 3rd time!
Sep 2025. Two papers will be presented on NeurIPS 2025 MLxOR Workshop. See you at San Diego on Dec 1-7!
Sep 2025. Our paper “Dynamic Pricing with Adversarially-Censored Demands” is accepted by WINE 2025! Thanks to my collaborators Yining, Xi and Yu-Xiang! The preprint version is posted here on ArXiv. See you at New Brunswick on Dec 7-10!
Jan 2025. Our manuscript “Joint Pricing and Resource Allocation: An Optimal Online-Learning Approach” is posted on ArXiv. Thanks to my collaborators Xuan, Yu-Xiang and Jiashuo!
Dec 2024. Glad to serve as an Area Chair in ICML 2025!