mlsanityData to 14 Oct 2026

EVE: A Generator-Verifier System for Generative Policies

arXiv 2512.21430 ↗First posted 24 Dec 2025cs.ROcited by 3

On the map: Diffusion models → Diffusion and flow policies for robotics

Abstract (arXiv)

Visuomotor policies based on generative such as diffusion and flow-matching have shown strong performance for robotics applications but degrade under distribution shifts, demonstrating limited recovery capabilities without costly finetuning. In the language modeling domain, test-time compute scaling has revolutionized the reasoning capabilities of modern LLMs by enabling candidate solution refinement. These methods typically leverage foundation models as verification modules in a zero-shot manner to score candidate solutions. We hypothesize that generative policies can similarly benefit from additional inference-time compute that employs zero-shot VLM-based verifiers in a generation-verification framework. To this end, we introduce EVE: a modular, generator-verifier interaction framework that boosts the performance of pretrained generative policies at test time, with no additional training. EVE wraps a frozen base policy with multiple zero-shot, VLM-based verifier agents. Each verifier proposes action refinements to the base policy candidate actions, while an action incorporator uses classifier guidance to fuse aggregated verifier feedback into action denoising. We study design choices for generator-verifier information interfacing across a system of verifiers with distinct capabilities. Across diverse simulated and real robotic tasks and embodiments, EVE consistently improves success rates without additional policy or verifier training. Through extensive ablations, we isolate the contribution of verifier capabilities and action incorporator strategies, offering practical guidelines to build scalable, modular generator-verifier systems for embodied control.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Placed in Diffusion and flow policies for robotics, one of 14 directions in Diffusion models. Builds on 49 of its 55 references that are in the map. First posted on arXiv 24 Dec 2025.

Builds on

49 of 55 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025Steering Vision-Language-Action Models as Anti-Exploration: A Test-Time Scaling Approach2 Dec
  2. 2025DynaGuide: Steering Diffusion Polices with Active Dynamic Guidance16 Jun
  3. 2024RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation10 Oct
  4. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusion7 Mar

Other references · 45

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  2. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  3. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,944
  4. 2022Flow Matching for Generative Modelingcited by 5,702
  5. 2023Let's Verify Step by Stepcited by 3,789
  6. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  7. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,407
  8. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,986
  9. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,761
  10. 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,225
  11. 2022Solving math word problems with process- and outcome-based feedbackcited by 901
  12. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 862
33 more
  1. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 852
  2. 2023VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Modelscited by 750
  3. 2023BridgeData V2: A Dataset for Robot Learning at Scalecited by 657
  4. 2025RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulationcited by 518
  5. 2024Evaluating Real-World Robot Manipulation Policies in Simulationcited by 495
  6. 2024Generative Verifiers: Reward Modeling as Next-Token Predictioncited by 480
  7. 2025AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systemscited by 405
  8. 2024Robotic Control via Embodied Chain-of-Thought Reasoningcited by 346
  9. 2025Real-Time Execution of Action Chunking Flow Policiescited by 216
  10. 2024PIVOT: Iterative Visual Prompting Elicits Actionable Knowledge for VLMscited by 190
  11. 2024MOKA: Open-World Robotic Manipulation through Mark-Based Visual Promptingcited by 129
  12. 2025A Careful Examination of Large Behavior Models for Multitask Dexterous Manipulationcited by 129
  13. 2024Steering Your Generalists: Improving Robotic Foundation Models via Value Guidancecited by 106
  14. 2024Vision Language Models are In-Context Value Learnerscited by 90
  15. 2025SAFE: Multitask Failure Detection for Vision-Language-Action Modelscited by 80
  16. 2025A Vision-Language-Action-Critic Model for Robotic Real-World Reinforcement Learningcited by 71
  17. 2024Unpacking Failure Modes of Generative Policies: Runtime Monitoring of Consistency and Progresscited by 69
  18. 2025Multi-Agent Verification: Scaling Test-Time Compute with Multiple Verifierscited by 62
  19. 2025RoboMonkey: Scaling Test-Time Sampling and Verification for Vision-Language-Action Modelscited by 61
  20. 2025Training Strategies for Efficient Embodied Reasoningcited by 55
  21. 2025From Foresight to Forethought: VLM-In-the-Loop Policy Steering via Latent Alignmentcited by 52
  22. 2024RACER: Rich Language-Guided Failure Recovery Policies for Imitation Learningcited by 49
  23. 2024Inference-Time Policy Steering through Human Interactionscited by 38
  24. 2024Deep Generative Models in Robotics: A Survey on Learning from Multimodal Demonstrationscited by 35
  25. 2025FailSafe: Reasoning and Recovery from Failures in Vision-Language-Action Modelscited by 35
  26. 2025Action-Free Reasoning for Policy Generalizationcited by 26
  27. 2025Latent Policy Barrier: Learning Robust Visuomotor Policies by Staying In-Distributioncited by 25
  28. 2024ManiSkill-HAB: A Benchmark for Low-Level Manipulation in Home Rearrangement Taskscited by 22
  29. 2025Robust Finetuning of Vision-Language-Action Robot Policies via Parameter Mergingcited by 22
  30. 2025FPC-VLA: A Vision-Language-Action Framework with a Supervisor for Failure Prediction and Correctioncited by 21
  31. 2025Variation in Verification: Understanding Verification Dynamics in Large Language Modelscited by 19
  32. 2025A Smooth Sea Never Made a Skilled SAILOR: Robust Imitation via Learning to Searchcited by 17
  33. 2025Learn from What We HAVE: History-Aware VErifier that Reasons about Past Interactions Onlinecited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Rewind-IL: Online Failure Detection and State Respawning for Imitation Learningcited by 9
  2. 2026Hide-and-Seek in Trajectories: Discovering Failure Signals for VLA Runtime Monitoringcited by 9
  3. 2026Flow-Support-Weighted Proposals for Test-Time Verified VLA Policiescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Looking Back to Move Forward: Temporal Verification for Generative Robot Policies
  2. 2026Visual Verification Enables Inference-time Steering and Autonomous Policy Improvement
  3. 2026Where Should Action Generation Begin? A Learnable Source Prior for Generative Robot Policies
  4. 2026Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment
  5. 2025Masked Generative Policy for Robotic Control
  6. 2025Inference-Time Enhancement of Generative Robot Policies via Predictive World Modeling
  7. 2025Much Ado About Noising: Dispelling the Myths of Generative Robotic Control
  8. 2026Retrieve-then-Steer: Online Success Memory for Test-Time Adaptation of Generative VLAs