Our paper “REFINE: Trajectory Representation Learning via Closed-Loop Transcription” was accepted at KDD 2026. Congratulations to Sean Bin Yang.