DiMaS: Distribution Matching for Steering Vision-Language-Action Models

Publication
Proceedings Track