Task Summary This dataset contains 20 episodes of teleoperated demonstrations with the SO-101 robot arm. The task is to move a white rook from a starting square to D4 on a real chessboard. 🎯 Goal square: D4 ♟️ Start squares: Episodes 0–9: A7 Episodes 10–19: B7 ⚠️ Note: Episode 5 is accidentally empty (skipped ahead during recording) The dataset was created at 30 FPS and includes full joint actions, proprioception, and a… See the full description on the dataset page: https://huggingface.co/datasets/7jep7/rook_to_d4_v5.