Minjae Kwon

Runtime Safety through Adaptive Shielding: From Hidden Parameter Inference to Provable Guarantees

Safe In-Context Reinforcement Learning

Logic-Based Adaptive Reward Shaping for Reinforcement Learning