Recovering Aggressively Pruned Vision-Language-Action Models with Offline Hidden-State Distillation

arXiv:2609.19579 · cs.RO · Submitted 2026-09-17 · Read on arXiv

cs.RO

Submitted: 2026-09-17

Updated: 2026-09-17

Code: https://github.com/microsoft/CogACT

Terminology

• Pruning — 58× in this paper · explained in 4 episodes
• Distillation — 24× in this paper · explained in 2 episodes
• Fine-Tuning — 11× in this paper · explained in 5 episodes
• Benchmark — 4× in this paper · explained in 2 episodes
• Transformer — 4× in this paper · explained in 2 episodes
• Reinforcement Learning — 3× in this paper · explained in 4 episodes
• Robustness — 3× in this paper · explained in 5 episodes
• Adaptation — 2× in this paper · explained in 1 episode
• Calibration — 2× in this paper · explained in 7 episodes
• Distribution Shift — 2× in this paper · explained in 5 episodes
• Action Decoder — 1× in this paper · explained in 1 episode
• Alignment — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 3 episodes
• Closed-Loop Control — 1× in this paper · explained in 2 episodes
• Embeddings — 1× in this paper · explained in 2 episodes
• Knowledge Distillation — 1× in this paper · explained in 11 episodes
• Redundancy — 1× in this paper · explained in 2 episodes
• VLA Models — 1× in this paper · explained in 3 episodes
• Vision-Language-Action (VLA) Models — 1× in this paper · explained in 2 episodes

Sources

• CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation
• Finetuning Vision-Language-Action Models Requires Fewer Layers Than You Think
• Drop-Then-Recovery: How Redundant Are Vision-Language-Action Models?
• A Survey on Efficient Vision-Language-Action Models
• The Better You Learn, The Smarter You Prune: Towards Efficient Vision-language-action Models via Differentiable Token Pruning
• SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics
• Don't Run with Scissors: Pruning Breaks VLA Models but They Can Be Recovered
• Revisiting Parameter Redundancy in Vision-Language-Action Models: Insights from VLM-to-VLA Adaptation
• Distilling the Knowledge in a Neural Network

Related papers

• FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
• MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
• RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
• HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
• APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies
• Fine-tuning is Not Enough: A Parallel Framework for Collaborative Imitation and Reinforcement Learning in End-to-end Autonomous Driving