# Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Models

arXiv: https://arxiv.org/abs/2505.21200 · HTML: https://mlsanity.com/paper/2505.21200/

## Key facts

- Cited by 30 papers in the map (data through 14 Oct 2026).
- Placed in Efficient VLA inference and adaptation, one of 13 directions in Vision-language-action models.
- Builds on 29 of its 48 references that are in the map.
- First posted on arXiv 27 May 2025.

## Placed in

- Vision-language-action models / [Efficient VLA inference and adaptation](https://mlsanity.com/vla/efficient-vla-inference-and-adaptation/)

## Builds on (most cited first, 29 of 29)

- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [OpenVLA: An Open-Source Vision-Language-Action Model](https://mlsanity.com/paper/2406.09246/) (2024)
- [Do As I Can, Not As I Say: Grounding Language in Robotic Affordances](https://mlsanity.com/paper/2204.01691/) (2022)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [RT-1: Robotics Transformer for Real-World Control at Scale](https://mlsanity.com/paper/2212.06817/) (2022)
- [Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success](https://mlsanity.com/paper/2502.19645/) (2025)
- [DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset](https://mlsanity.com/paper/2403.12945/) (2024)
- [R3M: A Universal Visual Representation for Robot Manipulation](https://mlsanity.com/paper/2203.12601/) (2022)
- [RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation](https://mlsanity.com/paper/2410.07864/) (2024)
- [FAST: Efficient Action Tokenization for Vision-Language-Action Models](https://mlsanity.com/paper/2501.09747/) (2025)
- [Evaluating Real-World Robot Manipulation Policies in Simulation](https://mlsanity.com/paper/2405.05941/) (2024)
- [CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation](https://mlsanity.com/paper/2411.19650/) (2024)
- [SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference](https://mlsanity.com/paper/2410.04417/) (2024)
- [Vision-Language Foundation Models as Effective Robot Imitators](https://mlsanity.com/paper/2311.01378/) (2023)
- [VisionZip: Longer is Better but Not Necessary in Vision Language Models](https://mlsanity.com/paper/2412.04467/) (2024)
- [GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulation](https://mlsanity.com/paper/2410.06158/) (2024)
- [RH20T: A Comprehensive Robotic Dataset for Learning Diverse Skills in One-Shot](https://mlsanity.com/paper/2307.00595/) (2023)
- [LLaVA-Mini: Efficient Image and Video Large Multimodal Models with One Vision Token](https://mlsanity.com/paper/2501.03895/) (2025)
- [RoboMamba: Efficient Vision-Language-Action Model for Robotic Reasoning and Manipulation](https://mlsanity.com/paper/2406.04339/) (2024)
- [PD-VLA: Accelerating Vision-Language-Action Model Integrated with Action Chunking via Parallel Decoding](https://mlsanity.com/paper/2503.02310/) (2025)
- [VLA-Cache: Efficient Vision-Language-Action Manipulation via Adaptive Token Caching](https://mlsanity.com/paper/2502.02175/) (2025)
- [Manipulate-Anything: Automating Real-World Robots using Vision-Language Models](https://mlsanity.com/paper/2406.18915/) (2024)
- [Multi-Stage Vision Token Dropping: Towards Efficient Multimodal Large Language Model](https://mlsanity.com/paper/2411.10803/) (2024)
- [Diffusion-VLA: Generalizable and Interpretable Robot Foundation Model via Self-Generated Reasoning](https://mlsanity.com/paper/2412.03293/) (2024)
- [Bidirectional Decoding: Improving Action Chunking via Guided Test-Time Sampling](https://mlsanity.com/paper/2408.17355/) (2024)
- [Diffusion Transformer Policy](https://mlsanity.com/paper/2410.15959/) (2024)
- [Inference Optimal VLMs Need Fewer Visual Tokens and More Parameters](https://mlsanity.com/paper/2411.03312/) (2024)
- [Quantization-Aware Imitation-Learning for Resource-Efficient Robotic Control](https://mlsanity.com/paper/2412.01034/) (2024)
- [DNAct: Diffusion Guided Multi-Task 3D Policy Learning](https://mlsanity.com/paper/2403.04115/) (2024)

## Built on by (most cited first, 30 of 30)

- [Large VLM-based Vision-Language-Action Models for Robotic Manipulation: A Survey](https://mlsanity.com/paper/2508.13073/) (2025)
- [The Better You Learn, The Smarter You Prune: Towards Efficient Vision-language-action Models via Differentiable Token Pruning](https://mlsanity.com/paper/2509.12594/) (2025)
- [Action-aware Dynamic Pruning for Efficient Vision-Language-Action Manipulation](https://mlsanity.com/paper/2509.22093/) (2025)
- [A Survey on Efficient Vision-Language-Action Models](https://mlsanity.com/paper/2510.24795/) (2025)
- [Efficient Vision-Language-Action Models for Embodied Manipulation: A Systematic Survey](https://mlsanity.com/paper/2510.17111/) (2025)
- [RLRC: Reinforcement Learning-based Recovery for Compressed Vision-Language-Action Models](https://mlsanity.com/paper/2506.17639/) (2025)
- [GeoSem-WAM: Geometry- and Semantic-Aware World Action Models](https://mlsanity.com/paper/2606.03188/) (2026)
- [AC^2-VLA: Action-Context-Aware Adaptive Computation in Vision-Language-Action Models for Efficient Robotic Manipulation](https://mlsanity.com/paper/2601.19634/) (2026)
- [ActDistill: General Action-Guided Self-Derived Distillation for Efficient Vision-Language-Action Models](https://mlsanity.com/paper/2511.18082/) (2025)
- [TIDAL: Temporally Interleaved Diffusion and Action Loop for High-Frequency VLA Control](https://mlsanity.com/paper/2601.14945/) (2026)
- [Efficient Long-Horizon Vision-Language-Action Models via Static-Dynamic Disentanglement](https://mlsanity.com/paper/2602.03983/) (2026)
- [OxyGen: Unified KV Cache Management for VLA Inference under Multi-Task Parallelism](https://mlsanity.com/paper/2603.14371/) (2026)
- [Revisiting Multimodal KV Cache Compression: A Frequency-Domain-Guided Outlier-KV-Aware Approach](https://mlsanity.com/paper/2511.16786/) (2025)
- [AttenA+: Rectifying Action Inequality in Robotic Foundation Models](https://mlsanity.com/paper/2605.13548/) (2026)
- [ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models](https://mlsanity.com/paper/2605.29438/) (2026)
- [Revisiting Parameter Redundancy in Vision-Language-Action Models: Insights from VLM-to-VLA Adaptation](https://mlsanity.com/paper/2606.31382/) (2026)
- [DepthCache: Depth-Guided Training-Free Visual Token Merging for Vision-Language-Action Model Inference](https://mlsanity.com/paper/2603.10469/) (2026)
- [Beyond Attention Magnitude: Leveraging Inter-layer Rank Consistency for Efficient Vision-Language-Action Models](https://mlsanity.com/paper/2603.24941/) (2026)
- [Overcoming Dynamics-Blindness: Training-Free Pace-and-Path Correction for VLA Models](https://mlsanity.com/paper/2605.11459/) (2026)
- [UniFS: Unified Fast-to-Slow Hierarchical Architecture for Vision-Language-Action Models](https://mlsanity.com/paper/2606.22794/) (2026)
- [Reducing Temporal Redundancy for Efficient Vision-Language-Action Inference](https://mlsanity.com/paper/2607.12287/) (2026)
- [UniMPA: A Unified Memory-Prediction-Action Model via Action-Grounded Transition Modeling](https://mlsanity.com/paper/2609.11875/) (2026)
- [VLA-InfoEntropy: A Training-Free Vision-Attention Information Entropy Approach for Vision-Language-Action Models Inference Acceleration and Success](https://mlsanity.com/paper/2604.05323/) (2026)
- [2D or 3D: Who Governs Salience in VLA Models? -- Tri-Stage Token Pruning Framework with Modality Salience Awareness](https://mlsanity.com/paper/2604.09244/) (2026)
- [Improving Robotic Imitation Learning via Trajectory Standardization](https://mlsanity.com/paper/2606.22907/) (2026)
- [FlashDrive: Flash Vision-Language-Action Inference for Autonomous Driving](https://mlsanity.com/paper/2608.12932/) (2026)
- [VLA-ULAP: Interleaving Cloud VLA Calls with Ultra-Lightweight Local Action Prediction at the Edge](https://mlsanity.com/paper/2609.18663/) (2026)
- [Catch Me If You Can: Real-Time Feedback Denoising for Responsive VLAs](https://mlsanity.com/paper/2609.21022/) (2026)
- [LLMs are General Asynchronous Agents](https://mlsanity.com/paper/2609.35427/) (2026)
- [VLA-ACL: Action-Consistent Visual Token Pruning for Efficient Vision-Language-Action Models](https://mlsanity.com/paper/2610.08133/) (2026)

## Abstract (arXiv)

Vision-Language-Action (VLA) models have emerged as a powerful paradigm for general-purpose robot control through natural language instructions. However, their high inference cost-stemming from large-scale token computation and autoregressive decoding-poses significant challenges for real-time deployment and edge applications. While prior work has primarily focused on architectural optimization, we take a different perspective by identifying a dual form of redundancy in VLA models: (i) high similarity across consecutive action steps, and (ii) substantial redundancy in visual tokens. Motivated by these observations, we propose FlashVLA, the first training-free and plug-and-play acceleration framework that enables action reuse in VLA models. FlashVLA improves inference efficiency through a token-aware action reuse mechanism that avoids redundant decoding across stable action steps, and an information-guided visual token selection strategy that prunes low-contribution tokens. Extensive experiments on the LIBERO benchmark show that FlashVLA reduces FLOPs by 55.7% and latency by 36.0%, with only a 0.7% drop in task success rate. These results demonstrate the effectiveness of FlashVLA in enabling lightweight, low-latency VLA inference without retraining.