# CARP: Visuomotor Policy Learning via Coarse-to-Fine Autoregressive Prediction

arXiv: https://arxiv.org/abs/2412.06782 · HTML: https://mlsanity.com/paper/2412.06782/

## Key facts

- Cited by 27 papers in the map (data through 14 Oct 2026).
- Builds on 66 of its 78 references that are in the map.
- First posted on arXiv 9 Dec 2024.

## Builds on (most cited first, 30 of 66)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [Neural Discrete Representation Learning](https://mlsanity.com/paper/1711.00937/) (2017)
- [Flamingo: a Visual Language Model for Few-Shot Learning](https://mlsanity.com/paper/2204.14198/) (2022)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Taming Transformers for High-Resolution Image Synthesis](https://mlsanity.com/paper/2012.09841/) (2020)
- [RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control](https://mlsanity.com/paper/2307.15818/) (2023)
- [OpenVLA: An Open-Source Vision-Language-Action Model](https://mlsanity.com/paper/2406.09246/) (2024)
- [Diffusion Policy: Visuomotor Policy Learning via Action Diffusion](https://mlsanity.com/paper/2303.04137/) (2023)
- [Do As I Can, Not As I Say: Grounding Language in Robotic Affordances](https://mlsanity.com/paper/2204.01691/) (2022)
- [Consistency Models](https://mlsanity.com/paper/2303.01469/) (2023)
- [Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware](https://mlsanity.com/paper/2304.13705/) (2023)
- [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://mlsanity.com/paper/2106.01345/) (2021)
- [Generating Diverse High-Fidelity Images with VQ-VAE-2](https://mlsanity.com/paper/1906.00446/) (2019)
- [Semantic Image Synthesis with Spatially-Adaptive Normalization](https://mlsanity.com/paper/1903.07291/) (2019)
- [On the Continuity of Rotation Representations in Neural Networks](https://mlsanity.com/paper/1812.07035/) (2018)
- [Planning with Diffusion for Flexible Behavior Synthesis](https://mlsanity.com/paper/2205.09991/) (2022)
- [Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction](https://mlsanity.com/paper/2404.02905/) (2024)
- [End to End Learning for Self-Driving Cars](https://mlsanity.com/paper/1604.07316/) (2016)
- [A Generalist Agent](https://mlsanity.com/paper/2205.06175/) (2022)
- [Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation](https://mlsanity.com/paper/2406.06525/) (2024)
- [What Matters in Learning from Offline Human Demonstrations for Robot Manipulation](https://mlsanity.com/paper/2108.03298/) (2021)
- [Autoregressive Image Generation using Residual Quantization](https://mlsanity.com/paper/2203.01941/) (2022)
- [Vector-quantized Image Modeling with Improved VQGAN](https://mlsanity.com/paper/2110.04627/) (2021)
- [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://mlsanity.com/paper/2106.02039/) (2021)
- [Is Conditional Generative Modeling all you need for Decision-Making?](https://mlsanity.com/paper/2211.15657/) (2022)

## Built on by (most cited first, 27 of 27)

- [OpenHelix: A Short Survey, Empirical Analysis, and Open-Source Dual-System VLA Model for Robotic Manipulation](https://mlsanity.com/paper/2505.03912/) (2025)
- [VLA-RFT: Vision-Language-Action Reinforcement Fine-tuning with Verified Rewards in World Simulators](https://mlsanity.com/paper/2510.00406/) (2025)
- [Long-VLA: Unleashing Long-Horizon Capability of Vision Language Action Model for Robot Manipulation](https://mlsanity.com/paper/2508.19958/) (2025)
- [VITA: Vision-to-Action Flow Matching Policy](https://mlsanity.com/paper/2507.13231/) (2025)
- [Dense Policy: Bidirectional Autoregressive Learning of Actions](https://mlsanity.com/paper/2503.13217/) (2025)
- [H3DP: Triply-Hierarchical Diffusion Policy for Visuomotor Learning](https://mlsanity.com/paper/2505.07819/) (2025)
- [Score and Distribution Matching Policy: Advanced Accelerated Visuomotor Policies via Matched Distillation](https://mlsanity.com/paper/2412.09265/) (2024)
- [Rethinking Latent Redundancy in Behavior Cloning: An Information Bottleneck Approach for Robot Manipulation](https://mlsanity.com/paper/2502.02853/) (2025)
- [FreqPolicy: Frequency Autoregressive Visuomotor Policy with Continuous Tokens](https://mlsanity.com/paper/2506.01583/) (2025)
- [CDP: Towards Robust Autoregressive Visuomotor Policy Learning via Causal Diffusion](https://mlsanity.com/paper/2506.14769/) (2025)
- [Mimic Intent, Not Just Trajectories](https://mlsanity.com/paper/2602.08602/) (2026)
- [SAMPO:Scale-wise Autoregression with Motion PrOmpt for generative world models](https://mlsanity.com/paper/2509.15536/) (2025)
- [Robotic Manipulation via Imitation Learning: Taxonomy, Evolution, Benchmark, and Challenges](https://mlsanity.com/paper/2508.17449/) (2025)
- [Fast Visuomotor Policy for Robotic Manipulation](https://mlsanity.com/paper/2510.12483/) (2025)
- [L1 Sample Flow for Efficient Visuomotor Learning](https://mlsanity.com/paper/2511.17898/) (2025)
- [R3DP: Real-Time 3D-Aware Policy for Embodied Manipulation](https://mlsanity.com/paper/2603.14498/) (2026)
- [HiFlow: Tokenization-Free Scale-Wise Autoregressive Policy Learning via Flow Matching](https://mlsanity.com/paper/2603.27281/) (2026)
- [CUBic: Coordinated Unified Bimanual Perception and Control Framework](https://mlsanity.com/paper/2605.13452/) (2026)
- [M-STAR: Multi-Scale Spatiotemporal Autoregression for Human Mobility Modeling](https://mlsanity.com/paper/2512.07314/) (2025)
- [ISS Policy : Scalable Diffusion Policy with Implicit Scene Supervision](https://mlsanity.com/paper/2512.15020/) (2025)
- [Referring-Aware Visuomotor Policy Learning for Closed-Loop Manipulation](https://mlsanity.com/paper/2604.05544/) (2026)
- [FocalPolicy: Frequency-Optimized Chunking and Locally Anchored Flow Matching for Coherent Visuomotor Policy](https://mlsanity.com/paper/2605.15944/) (2026)
- [B'ezier Degradation Modeling for LiDAR-based Human Motion Capture](https://mlsanity.com/paper/2605.19620/) (2026)
- [Robust Fuzzy Multi-view Learning under View Conflict](https://mlsanity.com/paper/2605.24475/) (2026)
- [ActionUNet: Improving Robustness of VLA Models with Efficient Multi-scale Fine-tuning](https://mlsanity.com/paper/2609.34982/) (2026)
- [Discrete Forcing: Infusing Discrete Guidance into Continuous Denoising for Few-Step Action Experts](https://mlsanity.com/paper/2609.39526/) (2026)
- [Hierarchical Residual Policy Learning for Real-World Mobile Manipulation with Sparse Human Guidance](https://mlsanity.com/paper/2610.20607/) (2026)

## Abstract (arXiv)

In robotic visuomotor policy learning, diffusion-based models have achieved significant success in improving the accuracy of action trajectory generation compared to traditional autoregressive models. However, they suffer from inefficiency due to multiple denoising steps and limited flexibility from complex constraints. In this paper, we introduce Coarse-to-Fine AutoRegressive Policy (CARP), a novel paradigm for visuomotor policy learning that redefines the autoregressive action generation process as a coarse-to-fine, next-scale approach. CARP decouples action generation into two stages: first, an action autoencoder learns multi-scale representations of the entire action sequence; then, a GPT-style transformer refines the sequence prediction through a coarse-to-fine autoregressive process. This straightforward and intuitive approach produces highly accurate and smooth actions, matching or even surpassing the performance of diffusion-based policies while maintaining efficiency on par with autoregressive policies. We conduct extensive evaluations across diverse settings, including single-task and multi-task scenarios on state-based and image-based simulation benchmarks, as well as real-world tasks. CARP achieves competitive success rates, with up to a 10% improvement, and delivers 10x faster inference compared to state-of-the-art policies, establishing a high-performance, efficient, and flexible paradigm for action generation in robotic tasks.