Xuening Feng

About

Xuening Feng is a rising researcher at the forefront of reinforcement learning, with a particular focus on advancing human-in-the-loop AI systems. Her key research areas include reinforcement learning from human feedback (RLHF), reward function design, and interactive machine learning. Feng’s most notable contribution is her work on the paper "DUO: Diverse, Uncertain, On-Policy Query Generation and Selection for Reinforcement Learning from Human Feedback" (2025), which addresses a critical bottleneck in RLHF: the challenge of efficiently generating and selecting informative queries for human feedback. By introducing a method that balances diversity, uncertainty, and on-policy relevance, Feng’s approach significantly improves the sample efficiency and robustness of RLHF systems, enabling agents to learn complex behaviors with less human effort. Though early in her career, her work has already garnered attention, with her flagship paper accumulating citations and sparking discussions in the AI community. Feng’s research holds promise for making RLHF more practical and scalable, paving the way for safer and more capable autonomous agents. Her innovative thinking and technical rigor mark her as a promising talent in the field.

Research Focus

Key Achievements

Top Papers

1. 1
   DUO: Diverse, Uncertain, On-Policy Query Generation and Selection for Reinforcement Learning from Human Feedback

   1 citations · 2025

Key Collaborators

Contact & Links