Do Better Goal Representations Improve Goal-Conditioned Reinforcement Learning?

Excerpt

arXiv:2609.39901v1 Announce Type: new Abstract: Goal-conditioned reinforcement learning (GCRL) relies heavily on how target goals are represented to the policy. While recent methods encode goals via temporal distance, occupancy, or controllability, it remains unclear how much downstream performance actually depends on representation quality. We study this in offline GCRL by constructing an exact temporal-distance goal representation in deterministic mazes. We then systematically corrupt its geom