Self-Predictive Representations for Combinatorial Generalization in Behavioral Cloning

arXiv:2506.10137v3 Announce Type: replace-cross Abstract: While goal-conditioned behavior cloning (GCBC) methods can perform well on in-distribution training tasks, they do not necessarily generalize zero-shot to tasks that require conditioning on novel state-goal pairs, i.e. combinatorial generalization. In part, this limitation can be attributed to a lack of temporal consistency in the state representation learned by BC; if temporally correlated states are properly encoded to similar latent re