2025 ICML ICML 2025

GoIRL: Graph-Oriented Inverse Reinforcement Learning for Multimodal Trajectory Prediction