This dataset contains simple robotics interaction data designed for
reinforcement learning and control experiments.
The dataset records state, action, and reward tuples collected from a
simulated mobile robot navigating a 2D environment.
episode index
timestep
robot position (x, y)
velocity
action taken
reward received
terminal state flag… See the full description on the dataset page:
https://huggingface.co/datasets/Goyoonjung/my-robotics-dataset-11.