mlsanityData to 7 Oct 2026

Hume: Introducing System-2 Thinking in Visual-Language-Action Model

arXiv 2505.21432 ↗First posted 27 May 2025cs.RO cs.AIcited by 56

On the map: LLM reasoning → Spatial and embodied reasoning

also in 1 more
Abstract (arXiv)

Humans practice slow thinking before performing actual actions when handling complex tasks in the physical world. This thinking paradigm, recently, has achieved remarkable advancement in boosting Large Language Models (LLMs) to solve complex tasks in digital domains. However, the potential of slow thinking remains largely unexplored for robotic foundation models interacting with the physical world. In this work, we propose Hume: a dual-system Vision-Language-Action (VLA) model with value-guided System-2 thinking and cascaded action denoising, exploring human-like thinking capabilities of Vision-Language-Action models for dexterous robot control. System 2 of Hume implements value-Guided thinking by extending a Vision-Language-Action Model backbone with a novel value-query head to estimate the state-action value of predicted actions. The value-guided thinking is conducted by repeat sampling multiple action candidates and selecting one according to state-action value. System 1 of Hume is a lightweight reactive visuomotor policy that takes System 2 selected action and performs cascaded action denoising for dexterous robot control. At deployment time, System 2 performs value-guided thinking at a low frequency while System 1 asynchronously receives the System 2 selected action candidate and predicts fluid actions in real time. We show that Hume outperforms the existing state-of-the-art Vision-Language-Action models across multiple simulation benchmark and real-robot deployments.

Key facts. Cited by 56 papers in the map (data through 7 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 65 of its 79 references that are in the map. First posted on arXiv 27 May 2025.

Builds on

65 of 79 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success27 Feb
  2. 2024COHERENT: Collaboration of Heterogeneous Multi-Robot System with Large Language Models23 Sep
  3. 2024Robotic Control via Embodied Chain-of-Thought Reasoning11 Jul

Other references · 62

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  2. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  3. 2022Flow Matching for Generative Modelingcited by 5,426
  4. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,064
  5. 2015Continuous control with deep reinforcement learningcited by 3,279
  6. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,155
  7. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
  8. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  9. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  10. 2018Addressing Function Approximation Error in Actor-Critic Methodscited by 2,302
  11. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  12. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,007
38 more
  1. 2020Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problemscited by 1,873
  2. 2024Octo: An Open-Source Generalist Robot Policycited by 1,427
  3. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  4. 2018Soft Actor-Critic Algorithms and Applicationscited by 1,179
  5. 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,167
  6. 2021Cascaded Diffusion Models for High Fidelity Image Generationcited by 1,155
  7. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 849
  8. 2023Reasoning with Language Model is Planning with World Modelcited by 846
  9. 2025FAST: Efficient Action Tokenization for Vision-Language-Action Modelscited by 643
  10. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Modelcited by 502
  11. 2024Evaluating Real-World Robot Manipulation Policies in Simulationcited by 484
  12. 2024CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulationcited by 411
  13. 2023Vision-Language Foundation Models as Effective Robot Imitatorscited by 344
  14. 2024Prismatic VLMs: Investigating the Design Space of Visually-Conditioned Language Modelscited by 336
  15. 2024TraceVLA: Visual Trace Prompting Enhances Spatial-Temporal Awareness for Generalist Robotic Policiescited by 298
  16. 2025O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruningcited by 278
  17. 2023PaLI-X: On Scaling up a Multilingual Vision and Language Modelcited by 254
  18. 2023Cal-QL: Calibrated Offline RL Pre-Training for Efficient Online Fine-Tuningcited by 230
  19. 2025Hi Robot: Open-Ended Instruction Following with Hierarchical Vision-Language-Action Modelscited by 216
  20. 2025DexVLA: Vision-Language Model with Plug-In Diffusion Expert for General Robot Controlcited by 191
  21. 2025FutureSightDrive: Thinking Visually with Spatio-Temporal CoT for Autonomous Drivingcited by 165
  22. 2023SelfCheck: Using LLMs to Zero-Shot Check Their Own Step-by-Step Reasoningcited by 157
  23. 2024Scaling Proprioceptive-Visual Learning with Heterogeneous Pre-trained Transformerscited by 145
  24. 2024What Matters in Building Vision-Language-Action Models for Generalist Robotscited by 88
  25. 2025Universal Actions for Enhanced Embodied Foundation Modelscited by 80
  26. 2022Pre-Training for Robots: Offline RL Enables Learning New Tasks from a Handful of Trialscited by 78
  27. 2024HiRT: Enhancing Robotic Control with Hierarchical Robot Transformerscited by 72
  28. 2024Towards Synergistic, Generalized, and Efficient Dual-System for Robotic Manipulationcited by 71
  29. 2024GroupDebate: Enhancing the Efficiency of Multi-Agent Debate Using Group Discussioncited by 64
  30. 2024Interpretable Contrastive Monte Carlo Tree Search Reasoningcited by 62
  31. 2025SETS: Leveraging Self-Verification and Self-Correction for Improved Test-Time Scalingcited by 54
  32. 2023Improving Large Language Model Fine-tuning for Solving Math Problemscited by 48
  33. 2024A Dual Process VLA: Efficient Robotic Manipulation Leveraging VLMcited by 44
  34. 2024FastUMI: A Scalable and Hardware-Independent Universal Manipulation Interface with Datasetcited by 36
  35. 2024MLLMGuard: A Multi-dimensional Safety Evaluation Suite for Multimodal Large Language Modelscited by 34
  36. 2022f-DM: A Multi-stage Diffusion Model via Progressive Signal Transformationcited by 32
  37. 2025Spatial-Temporal Graph Diffusion Policy with Kinematic Modeling for Bimanual Robotic Manipulationcited by 24
  38. 2024Learning 2D Invariant Affordance Knowledge for 3D Affordance Groundingcited by 21

50 most cited of 62 All 62 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 24; 2026: 32.

56 papers.

Sort:

By area:

  1. 2025DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledgecited by 203
  2. 2025VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Modelcited by 140
  3. 2025Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policiescited by 90
  4. 2025F1: A Vision-Language-Action Model Bridging Understanding and Generation to Actionscited by 85
  5. 2025InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policycited by 83
  6. 2025Vision-Language-Action Models for Robotics: A Review Towards Real-World Applicationscited by 82
  7. 2025Pure Vision Language Action (VLA) Models: A Comprehensive Surveycited by 46
  8. 2025CO-RFT: Efficient Fine-Tuning of Vision-Language-Action Models through Chunked Offline Reinforcement Learningcited by 34
  9. 2025EO-1: An Open Unified Embodied Foundation Model for General Robot Controlcited by 30
  10. 2025Efficient Vision-Language-Action Models for Embodied Manipulation: A Systematic Surveycited by 25
  11. 2026Adaptive Action Chunking at Inference-time for Vision-Language-Action Modelscited by 24
  12. 2026How Fast Can I Run My VLA? Demystifying VLA Inference Performance with VLA-Perfcited by 17
38 more
  1. 2025RoVer: Robot Reward Model as Test-Time Verifier for Vision-Language-Action Modelcited by 16
  2. 2025RobustVLA: Robustness-Aware Reinforcement Post-Training for Vision-Language-Action Modelscited by 16
  3. 2025A Step Toward World Models: A Survey on Robotic Manipulationcited by 15
  4. 2025OmniDrive-R1: Reinforcement-driven Interleaved Multi-modal Chain-of-Thought for Trustworthy Vision-Language Autonomous Drivingcited by 9
  5. 2026VLA-ATTC: Adaptive Test-Time Compute for VLA Models with Relative Action Critic Modelcited by 9
  6. 2025Towards Deploying VLA without Fine-Tuning: Plug-and-Play Inference-Time VLA Policy Steering via Embodied Evolutionary Diffusioncited by 8
  7. 2026Omni2Sound: Towards Unified Video-Text-to-Audio Generationcited by 8
  8. 2026ST4VLA: Spatially Guided Training for Vision-Language-Action Modelscited by 8
  9. 2025Survey of Vision-Language-Action Models for Embodied Manipulationcited by 7
  10. 2025RoboOmni: Proactive Robot Manipulation in Omni-modal Contextcited by 7
  11. 2025Robotic VLA Benefits from Joint Learning with Motion Image Diffusioncited by 7
  12. 2025Asynchronous Fast-Slow Vision-Language-Action Policies for Whole-Body Robotic Manipulationcited by 6
  13. 2026AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Modelscited by 6
  14. 2026π-StepNFT: Wider Space Needs Finer Steps in Online RL for Flow-based VLAscited by 5
  15. 2026ZeroWBC: Learning Natural Whole-Body Humanoid Interaction from Human Egocentric Datacited by 5
  16. 2026When Models Learn to Ask Why: Adaptive Causal Reasoning for Trustworthy Medical Vision-Language Modelscited by 5
  17. 2025Progressive Gaussian Transformer with Anisotropy-aware Sampling for Open Vocabulary Occupancy Predictioncited by 4
  18. 2025An Anatomy of Vision-Language-Action Models: From Modules to Milestones and Challengescited by 4
  19. 2026Fine-tuning is Not Enough: A Parallel Framework for Collaborative Imitation and Reinforcement Learning in End-to-end Autonomous Drivingcited by 4
  20. 2026UAV-Track VLA: Embodied Aerial Tracking via Vision-Language-Action Modelscited by 4
  21. 2026Sentinel-VLA: A Metacognitive VLA Model with Active Status Monitoring for Dynamic Reasoning and Error Recoverycited by 4
  22. 2026TDEC: Deep Embedded Image Clustering with Transformer and Distribution Informationcited by 3
  23. 2026Hyper-DP3: Frequency-Aware Right-Sizing of 3D Diffusion Policies for Visuomotor Controlcited by 3
  24. 2025Trajectory Conditioned Cross-embodiment Skill Transfercited by 2
  25. 2025Language-Conditioned Representations and Mixture-of-Experts Policy for Robust Multi-Task Robotic Manipulationcited by 2
  26. 2025Beyond Success: Refining Elegant Robot Manipulation from Mixed-Quality Data via Just-in-Time Interventioncited by 2
  27. 2026GeodesicNVS: Probability Density Geodesic Flow Matching for Novel View Synthesiscited by 2
  28. 2026PhyGile: Physics-Prefix Guided Motion Generation for Agile General Humanoid Motion Trackingcited by 2
  29. 2026From Imagined Futures to Executable Actions: Mixture of Latent Actions for Robot Manipulationcited by 2
  30. 2026Partially Observable Adversarial Patch Attacks on Vision-Language-Action Models in Roboticscited by 2
  31. 2026Q-VGM: Q-Guided Value-Gradient Matching for Offline-to-Online RL of Flow-Matching VLA Policiescited by 2
  32. 2026FontCrafter: High-Fidelity Element-Driven Artistic Font Creation with Visual In-Context Generationcited by 1
  33. 2026CNMBI: Determining the Number of Clusters Using Center Pairwise Matching and Boundary Filteringcited by 1
  34. 2026Deep Image Clustering Based on Curriculum Learning and Density Informationcited by 1
  35. 2026UniFS: Unified Fast-to-Slow Hierarchical Architecture for Vision-Language-Action Modelscited by 1
  36. 2026E-TTS: A New Embodied Test-Time Scaling Framework for Robotic Manipulationcited by 1
  37. 2026From Sparse Decisions to Dense Reasoning: A Multi-attribute Trajectory Paradigm for Multimodal Moderationcited by 0
  38. 2026LoD-Loc v3: Generalized Aerial Localization in Dense Cities using Instance Silhouette Alignmentcited by 0

50 most cited of 56 All 56 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026AC^2-VLA: Action-Context-Aware Adaptive Computation in Vision-Language-Action Models for Efficient Robotic Manipulation
  2. 2025Fast-in-Slow: A Dual-System Foundation Model Unifying Fast Manipulation within Slow Reasoning
  3. 2026ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models
  4. 2025Large VLM-based Vision-Language-Action Models for Robotic Manipulation: A Survey
  5. 2026VLA-Thinker: Boosting Vision-Language-Action Models through Thinking-with-Image Reasoning
  6. 2025AVA-VLA: Improving Vision-Language-Action models with Active Visual Attention
  7. 2026DAM-VLA: A Dynamic Action Model-Based Vision-Language-Action Framework for Robot Manipulation
  8. 2024Language-Conditioned Robotic Manipulation with Fast and Slow Thinking