Jinfeng Li

Papers

Total Citations

H-Index

About

Jinfeng Li is a researcher advancing the frontiers of deep reinforcement learning (RL), with a particular focus on integrating human feedback into autonomous decision-making systems. Their most cited work, "Offline reward shaping with scaling human preference feedback for deep reinforcement learning" (2024), tackles a critical challenge in RL: aligning agent behavior with complex human values without requiring costly online interactions. By developing a framework that leverages offline datasets and scalable human preference signals, Li enables more efficient and ethically grounded reward shaping—a contribution that bridges the gap between theoretical RL and real-world deployment. Though early in their career, this work has already garnered 4 citations, signaling growing interest from peers in human-in-the-loop AI. Li’s research sits at the intersection of reinforcement learning, human-computer interaction, and algorithmic fairness, offering practical pathways for training agents that better reflect human intent. Their approach not only reduces the computational burden of traditional reward design but also opens doors for safer, more interpretable AI systems. As the demand for aligned AI grows, Li’s innovations in offline preference learning position them as a rising voice in shaping how machines learn from human guidance.