Recoleta
GitHub
Source note

PriorVLA: Prior-Preserving Adaptation for Vision-Language-Action Models

Authors
Xinyu Guo, Bin Xie, Wei Chai, Xianchi Deng, Tiancai Wang, Zhengxing Wu, et al.
Source
arXiv
Published
Collection
Embodied AI
Vision Language ActionRobot Foundation ModelGeneralist Robot PolicyVLA AdaptationRobot Data Scaling
Open arXivMarkdown source

Summary

PriorVLA adapts pretrained VLA robot policies without overwriting useful pretraining priors. It keeps a frozen action expert as a prior source and trains a separate adaptation expert, with the largest gains in OOD and few-shot robot manipulation.

Problem

  • Full fine-tuning of pretrained VLAs can overfit limited downstream robot data and reduce OOD generalization.
  • This matters because new robot tasks, scenes, objects, and embodiments often have few demonstrations, so losing pretrained scene and motor priors wastes costly large-scale pretraining.

Approach

  • PriorVLA starts from pi0.5 and splits the action model into two paths: a frozen Prior Expert and a trainable Adaptation Expert initialized from the same weights.
  • The frozen Prior Expert runs during denoising, but its final action prediction is discarded; its internal states provide motor-prior features.
  • Learnable Scene Queries read task-relevant visual-language features from the VLM, Motor Queries read frozen action-denoising features, and Action Queries feed both into the Adaptation Expert.
  • Training uses the same flow-matching MSE action loss as full fine-tuning, applied only to the Adaptation Expert.
  • The method trains the Adaptation Expert, Expert Queries, and VLM vision encoder while freezing the rest, updating about 25% as many parameters as full fine-tuning.

Results

  • On RoboTwin 2.0 standard training across 13 tasks, PriorVLA reaches 77% Easy ID and 53% Hard OOD success, beating pi0.5 by +10 and +11 points.
  • On RoboTwin 2.0 data scaling, Hard OOD gains over pi0.5 are +11 points in few-shot, +11 in standard data, and +6 in large-data settings; Easy ID is 41% vs 29% in few-shot and 77% vs 67% in standard data.
  • On LIBERO, PriorVLA reports 99.1% average success across Spatial, Object, Goal, and Long, above OpenVLA-OFT at 97.1% and pi0.5 at 96.9%.
  • In real-world standard-data tests over eight tasks and two robot embodiments, PriorVLA reaches 81% ID and 57% OOD success, improving over pi0.5 by +12 and +16 points.
  • With 10 real-world demonstrations per task, PriorVLA reaches 48% ID and 32% OOD success, beating pi0.5 by +24 and +22 points.

Link