Shimin Zhang

Training datasets and released checkpoints for RLVR from the CVPR 2026 paper “ReLaX: Reasoning with Latent Exploration for Large Reasoning Models”