How Many Demonstrations Does a VLA Need? On the Sample Efficiency of VLA Finetuning
cs.CV, cs.RO
Submitted: 2026-10-28
Updated: 2026-11-10
Code: https://github.com/NVIDIA/Isaac-GR00T
Project page: https://mapsvla.github.io
Terminology
Sources
- $\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization
- GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
- On the Efficiency of LoRA Fine-Tuning for Vision-Language-Action Models in Industrial Robotic Manipulation
- FOCA: Future-Oriented Conditioning for Data-Efficient Vision-Language-Action Adaptation
- MoS-VLA: A Vision-Language-Action Model with One-Shot Skill Adaptation
- Don't Blind Your VLA: Aligning Visual Representations for OOD Generalization
- VLM4VLA: Revisiting Vision-Language-Models in Vision-Language-Action Models
- Overtrained Language Models Are Harder to Fine-Tune
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs