VLA-ACL: Action-Consistent Visual Token Pruning for Efficient Vision-Language-Action Models
cs.RO, cs.CV
Submitted: 2026-10-06
Updated: 2026-10-08
Terminology
Sources
- $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
- $\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization
- Bridging the Semantic-Action Gap in Visual Token Pruning for Efficient VLA Inference
- Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Models
- The Better You Learn, The Smarter You Prune: Towards Efficient Vision-language-action Models via Differentiable Token Pruning
- PaliGemma: A versatile 3B VLM for transfer
- Llama 2: Open Foundation and Fine-Tuned Chat Models
- SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics
- The Gate, Not the Cache: Gate Provenance Bounds the Closed-Loop Reliability of Training-Free VLA Token Skipping
- Beyond Surrogate Gradients: Fully Differentiable Token Pruning for Vision-Language Models
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
- ARC: A Reasoning Recipe for Robot Foundation Models
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies