Quan Xiao

Ph.D. student in
Department of Electrical and Computer Engineering
Cornell University

quanx1808@gmail.com

Address: 2 W Loop Rd, New York, NY, 10044

Blog

Sep 29, 2026
BRIDGE: Bilevel Retrieval-Credit-Aware Agentic Reinforcement Learning

BRIDGE trains an LLM agent and its retriever through bilevel optimization, using feedback from the agent's answers to improve the evidence available for reasoning.

Read article