mlsanityData to 7 Oct 2026

FASTer: Toward Efficient Autoregressive Vision Language Action Modeling via Neural Action Tokenization

arXiv 2512.04952 ↗First posted 4 Dec 2025cs.CV cs.ROcited by 21

On the map: Vision-language-action models → Efficient VLA inference and adaptation

Abstract (arXiv)

Autoregressive vision-language-action (VLA) models have recently demonstrated strong capabilities in robotic manipulation. However, their core process of action tokenization often involves a trade-off between reconstruction fidelity and inference efficiency. We introduce FASTer, a unified framework for efficient and generalizable robot learning that integrates a learnable tokenizer with an autoregressive policy built upon it. FASTerVQ encodes action chunks as single-channel images, capturing global spatio-temporal dependencies while maintaining a high compression ratio. FASTerVLA builds on this tokenizer with block-wise autoregressive decoding and a lightweight action expert, achieving both faster inference and higher task performance. Extensive experiments across simulated and real-world benchmarks show that FASTerVQ delivers superior reconstruction quality, high token utilization, and strong cross-task and cross-embodiment generalization, while FASTerVLA further improves overall capability, surpassing previous state-of-the-art VLA models in both inference speed and task performance.

Key facts. Cited by 21 papers in the map (data through 7 Oct 2026). Placed in Efficient VLA inference and adaptation, one of 13 directions in Vision-language-action models. Builds on 61 of its 67 references that are in the map. First posted on arXiv 4 Dec 2025.

Builds on

61 of 67 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025VQ-VLA: Improving Vision-Language-Action Models via Scaling Vector-Quantized Action Tokenizers1 Jul
  2. 2025Conditioning Matters: Training Diffusion Policies is Faster Than You Think16 May
  3. 2025FAST: Efficient Action Tokenization for Vision-Language-Action Models16 Jan

Other references · 58

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2024The Llama 3 Herd of Modelscited by 14,192
  3. 2023Visual Instruction Tuningcited by 7,792
  4. 2024Qwen2.5 Technical Reportcited by 4,713
  5. 2017Neural Discrete Representation Learningcited by 4,650
  6. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
  7. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  8. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  9. 2021BEiT: BERT Pre-Training of Image Transformerscited by 2,360
  10. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  11. 2023PaLM-E: An Embodied Multimodal Language Modelcited by 2,037
  12. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,653
38 more
  1. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  2. 2024Octo: An Open-Source Generalist Robot Policycited by 1,427
  3. 2025InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiencycited by 1,419
  4. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  5. 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,167
  6. 2024Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Predictioncited by 1,005
  7. 2025Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Successcited by 844
  8. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 801
  9. 2018QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulationcited by 730
  10. 2022Autoregressive Image Generation using Residual Quantizationcited by 706
  11. 2023BridgeData V2: A Dataset for Robot Learning at Scalecited by 634
  12. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Modelcited by 502
  13. 2024Evaluating Real-World Robot Manipulation Policies in Simulationcited by 484
  14. 2025UniVLA: Learning to Act Anywhere with Task-centric Latent Actionscited by 386
  15. 2025Block Diffusion: Interpolating Between Autoregressive and Diffusion Language Modelscited by 385
  16. 2024Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robotscited by 373
  17. 2022MAGVIT: Masked Generative Video Transformercited by 341
  18. 2024An Image is Worth 32 Tokens for Reconstruction and Generationcited by 295
  19. 2025LongLive: Real-time Interactive Long Video Generationcited by 237
  20. 2024PaliGemma 2: A Family of Versatile VLMs for Transfercited by 204
  21. 2024VLABench: A Large-Scale Benchmark for Language-Conditioned Robotics Manipulation with Long-Horizon Reasoning Taskscited by 171
  22. 2022More ConvNets in the 2020s: Scaling up Kernels Beyond 51x51 using Sparsitycited by 142
  23. 2024OmniTokenizer: A Joint Image-Video Tokenizer for Visual Generationcited by 117
  24. 2024Image and Video Tokenization with Binary Spherical Quantizationcited by 110
  25. 2025Galaxea Open-World Dataset and G0 Dual-System VLA Modelcited by 90
  26. 2024WorldDreamer: Towards General World Models for Video Generation via Predicting Masked Tokenscited by 87
  27. 2023SpeechTokenizer: Unified Speech Tokenizer for Speech Large Language Modelscited by 85
  28. 2025Actions as Language: Fine-Tuning VLMs into VLAs Without Catastrophic Forgettingcited by 52
  29. 2024Scaling Transformers for Low-Bitrate High-Quality Speech Codingcited by 48
  30. 2024VidTok: A Versatile and Open-Source Video Tokenizercited by 35
  31. 2025Flow to the Mode: Mode-Seeking Diffusion Autoencoders for State-of-the-Art Image Tokenizationcited by 35
  32. 2025VLA-OS: Structuring and Dissecting Planning Representations and Paradigms in Vision-Language-Action Modelscited by 34
  33. 2024Autoregressive Action Sequence Learning for Robotic Manipulationcited by 26
  34. 2025BEAST: Efficient Tokenization of B-Splines Encoded Action Sequences for Imitation Learningcited by 25
  35. 2025XY-Tokenizer: Mitigating the Semantic-Acoustic Conflict in Low-Bitrate Speech Codecscited by 21
  36. 2024Learning Graph Quantized Tokenizerscited by 19
  37. 2024Low Frame-rate Speech Codec: a Codec Designed for Fast High-quality Speech LLM Training and Inferencecited by 11
  38. 2024GLAD: Improving Latent Graph Generative Modeling with Simple Quantizationcited by 8

50 most cited of 58 All 58 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 20.

21 papers.

Sort:

By area:

  1. 2026CLAP: Contrastive Latent Action Pretraining for Learning Vision-Language-Action Models from Human Videoscited by 27
  2. 2026LAP: Language-Action Pre-Training Enables Zero-shot Cross-Embodiment Transfercited by 17
  3. 2026ActionCodec: What Makes for Good Action Tokenizerscited by 17
  4. 2026Wall-OSS-0.5 Technical Reportcited by 11
  5. 2025Unified Embodied VLM Reasoning with Robotic Action via Autoregressive Discretized Pre-trainingcited by 7
  6. 2026AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Modelscited by 6
  7. 2026X-Tokenizer: A Multimodal Action Tokenizer for Vision-Language-Action Pretrainingcited by 6
  8. 2026Mimic Intent, Not Just Trajectoriescited by 5
  9. 2026DFM-VLA: Iterative Action Refinement for Robot Manipulation via Discrete Flow Matchingcited by 5
  10. 2026Revisiting Embodied Chain-of-Thought for Generalizable Robot Manipulationcited by 4
  11. 2026Coarse-to-Control: Action-Token Planning for Vision-Language-Action Modelscited by 3
  12. 2026Reflex: Real-Time VLA Control through Streaming Inferencecited by 3
9 more
  1. 2026Robots Need More than VLA and World Modelscited by 2
  2. 2026Towards Predictive, Aligned, and Scalable Robot Learningcited by 2
  3. 2026Overcoming Dynamics-Blindness: Training-Free Pace-and-Path Correction for VLA Modelscited by 1
  4. 2026Real-Time Execution with Autoregressive Policiescited by 0
  5. 2026ActionPiece: Rethinking Action Tokenization for Autoregressive Vision-Language-Action Modelscited by 0
  6. 2026CARE: Experience-Guided Atomic Corrective Execution for Vision-Language-Action Policiescited by 0
  7. 2026Rethinking Causal Action Tokenization with Conditional Annealing in Flow Matchingcited by 0
  8. 2026Humanoid Loco-Manipulation With Discrete VLA Modelcited by 0
  9. 2026SplineWAM: Adaptive Action Horizons for World Action Models via B-Spline Representationscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Models
  2. 2025PD-VLA: Accelerating Vision-Language-Action Model Integrated with Action Chunking via Parallel Decoding
  3. 2026Reducing Temporal Redundancy for Efficient Vision-Language-Action Inference
  4. 2026G0.5: One Autoregressive Stream for Robot Reasoning and Action
  5. 2026Learning to Accelerate Vision-Language-Action Models through Adaptive Visual Token Caching
  6. 2026A1: A Fully Transparent Open-Source, Adaptive and Efficient Truncated Vision-Language-Action Model
  7. 2026FlashVLA: Streaming Action Decoding for Fast and Asynchronous VLA Inference
  8. 2026LoopVLA: Learning Sufficiency in Recurrent Refinement for Vision-Language-Action Models