Observations Meet Actions: Learning Control-Sufficient Representations for Robust Policy Generalization

Open in new window