PermVLA: Factorization Order as a Regularizer for VLA Learning

Excerpt

arXiv:2610.04659v1 Announce Type: new Abstract: Vision-language-action (VLA) policies commonly learn action chunks through a fixed left-to-right (LTR) factorization, although the same expert trajectory distribution admits many valid chain-rule factorizations. We identify factorization order as an overlooked regularization choice and introduce causally anchored permutation (CAP), which samples action reveal orders with a tunable chronological prefix. Its auxiliary objective trains one shared poli