ACG: Action Coherence Guidance for Flow-based Vision-Language-Action models
On the map: Vision-language-action models → Efficient VLA inference and adaptation
also in 1 more
Abstract (arXiv)
Diffusion and flow matching models have emerged as powerful robot policies, enabling Vision-Language-Action (VLA) models to generalize across diverse scenes and instructions. Yet, when trained via imitation learning, their high generative capacity makes them sensitive to noise in human demonstrations: jerks, pauses, and jitter which reduce action coherence. Reduced action coherence causes instability and trajectory drift during deployment, failures that are catastrophic in fine-grained manipulation where precision is crucial. In this paper, we present Action Coherence Guidance (ACG) for VLA models, a training-free test-time guidance algorithm that improves action coherence and thereby yields performance gains. Evaluated on RoboCasa, DexMimicGen, and real-world SO-101 tasks, ACG consistently improves action coherence and boosts success rates across diverse manipulation tasks. Code and project page are available at https://github.com/DAVIAN-Robotics/ACG and https://DAVIAN-Robotics.github.io/ACG , respectively.
Key facts. Cited by 1 paper 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 43 of its 51 references that are in the map. First posted on arXiv 25 Oct 2025.
Builds on
43 of 51 references.
Closest ancestors · 1
Same direction, newest first
- 2025HybridVLA: Collaborative Diffusion and Autoregression in a Unified Vision-Language-Action Model13 Mar
Other references · 42
- 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
- 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
- 2022Scalable Diffusion Models with Transformerscited by 5,931
- 2022Classifier-Free Diffusion Guidancecited by 5,717
- 2022Flow Matching for Generative Modelingcited by 5,426
- 2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 3,550
- 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
- 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
- 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
- 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
- 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,007
- 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
30 more
- 2024Octo: An Open-Source Generalist Robot Policycited by 1,427
- 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,167
- 2021What Matters in Learning from Offline Human Demonstrations for Robot Manipulationcited by 699
- 2024RDT-1B: a Diffusion Foundation Model for Bimanual Manipulationcited by 670
- 2025SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Roboticscited by 507
- 2021Implicit Behavioral Cloningcited by 442
- 2024RoboCasa: Large-Scale Simulation of Everyday Tasks for Generalist Robotscited by 391
- 2022Behavior Transformers: Cloning k modes with one stonecited by 299
- 2017Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperationcited by 286
- 2024Guiding a Diffusion Model with a Bad Version of Itselfcited by 286
- 2023Imitating Human Behaviour with Diffusion Modelscited by 260
- 2023Goal-Conditioned Imitation Learning using Score-based Diffusion Policiescited by 229
- 2025Real-Time Execution of Action Chunking Flow Policiescited by 199
- 2025DreamGen: Unlocking Generalization in Robot Learning through Video World Modelscited by 168
- 2022Improving Sample Quality of Diffusion Models Using Self-Attention Guidancecited by 151
- 2024Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidancecited by 108
- 2019Self-Supervised Correspondence in Visuomotor Policy Learningcited by 106
- 2024DexMimicGen: Automated Data Generation for Bimanual Dexterous Manipulation via Imitation Learningcited by 104
- 2023Guided Flows for Generative Modeling and Decision Makingcited by 99
- 2023Waypoint-Based Imitation Learning for Robotic Manipulationcited by 59
- 2024Smoothed Energy Guidance: Guiding Diffusion Models with Reduced Energy Curvature of Attentioncited by 55
- 2023HYDRA: Hybrid Robot Actions for Imitation Learningcited by 54
- 2025Diffusion Guidance Is a Controllable Policy Improvement Operatorcited by 51
- 2024Bidirectional Decoding: Improving Action Chunking via Guided Test-Time Samplingcited by 46
- 2024Spatiotemporal Skip Guidance for Enhanced Video Diffusion Samplingcited by 33
- 2020Robust Imitation Learning from Noisy Demonstrationscited by 26
- 2023Good Better Best: Self-Motivated Imitation Learning for noisy Demonstrationscited by 4
- 2025LiPo: A Lightweight Post-optimization Framework for Smoothing Action Chunks Generated by Learned Policiescited by 4
- 2025Demonstration Sidetracks: Categorizing Systematic Non-Optimality in Human Demonstrationscited by 2
- 2025Self-Guided Action Diffusioncited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.
Similar, not cited
Close in wording, no citation either way
- 2026Self-Improving VLA Policies: Selected Diffusion Noise for Spurious-Robust Action Smoothing
- 2024VLATest: Testing and Evaluating Vision-Language-Action Models for Robotic Manipulation
- 2024CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation
- 2026Action ControlNet: A Lightweight Delay-Aware Adapter for Smooth Asynchronous Control in Vision-Language-Action Models
- 2026When to Switch: Reliable Action-Chunk Extension for Vision-Language-Action Models
- 2026A1: A Fully Transparent Open-Source, Adaptive and Efficient Truncated Vision-Language-Action Model
- 2026Catch Me If You Can: Real-Time Feedback Denoising for Responsive VLAs
- 2026CofactVLA: Deconfounding Vision-Language-Action Models via Counterfactual Intervention