mlsanityData to 14 Oct 2026

ReWiND: Language-Guided Rewards Teach Robot Policies without New Demonstrations

arXiv 2505.10911 ↗First posted 16 May 2025cs.ROcited by 60

On the map: Offline RL and imitation learning from data → Language-conditioned policies and VLAs

Abstract (arXiv)

We introduce ReWiND, a framework for learning robot manipulation tasks solely from language instructions without per-task demonstrations. Standard reinforcement learning (RL) and imitation learning methods require expert supervision through human-designed reward functions or demonstrations for every new task. In contrast, ReWiND starts from a small demonstration dataset to learn: (1) a data-efficient, language-conditioned reward function that labels the dataset with rewards, and (2) a language-conditioned policy pre-trained with offline RL using these rewards. Given an unseen task variation, ReWiND fine-tunes the pre-trained policy using the learned reward function, requiring minimal online interaction. We show that ReWiND's reward model generalizes effectively to unseen tasks, outperforming baselines by up to 2.4x in reward generalization and policy alignment metrics. Finally, we demonstrate that ReWiND enables sample-efficient adaptation to new tasks, beating baselines by 2x in simulation and improving real-world pretrained bimanual policies by 5x, taking a step towards scalable, real-world robot learning. See website at https://rewind-reward.github.io/.

Key facts. Cited by 60 papers in the map (data through 14 Oct 2026). Placed in Language-conditioned policies and VLAs, one of 13 directions in Offline RL and imitation learning from data. First posted on arXiv 16 May 2025.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 13; 2026: 47.

60 papers.

Sort:

By area:

  1. 2025π*0.6: a VLA That Learns From Experiencecited by 303
  2. 2025Alpamayo-R1: Bridging Reasoning and Action Prediction for Generalizable Autonomous Driving in the Long Tailcited by 146
  3. 2025Large VLM-based Vision-Language-Action Models for Robotic Manipulation: A Surveycited by 69
  4. 2026Robometer: Scaling General-Purpose Robotic Reward Models via Trajectory Comparisonscited by 60
  5. 2025Pure Vision Language Action (VLA) Models: A Comprehensive Surveycited by 49
  6. 2026TOPReward: Token Probabilities as Hidden Zero-Shot Rewards for Roboticscited by 33
  7. 2025EVOLVE-VLA: Test-Time Training from Environment Feedback for Vision-Language-Action Modelscited by 22
  8. 2025RobustVLA: Robustness-Aware Reinforcement Post-Training for Vision-Language-Action Modelscited by 18
  9. 2026Towards Long-Lived Robots: Continual Learning VLA Models via Reinforcement Fine-Tuningcited by 13
  10. 2026chi0: Resource-Aware Robust Manipulation via Taming Distributional Inconsistenciescited by 12
  11. 2026ARM: Advantage Reward Modeling for Long-Horizon Manipulationcited by 10
  12. 2025OpenGVL -- Benchmarking Visual Temporal Progress for Data Curationcited by 9
38 more
  1. 2026Beyond Imitation: Reinforcement Learning-Based Sim-Real Co-Training for VLA Modelscited by 9
  2. 2026ProcVLM: Learning Procedure-Grounded Progress Rewards for Robotic Manipulationcited by 9
  3. 2026RynnValue: Scaling Robotic Value Foundation Models with Temporal Distancecited by 7
  4. 2026From Passive Observer to Active Critic: Reinforcement Learning Elicits Process Reasoning for Robotic Manipulationcited by 6
  5. 2026Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptationcited by 6
  6. 2026TapSampling: Inference-Time Sampling with a Task-Progress-Understanding Verifier for Robotic Manipulationcited by 5
  7. 2026Progress Reward Modeling for Robotic Learning: A Comprehensive Surveycited by 5
  8. 2025AutoFocus-IL: VLM-based Saliency Maps for Data-Efficient Visual Imitation Learning without Extra Human Annotationscited by 4
  9. 2025OMP: One-step Meanflow Policy with Directional Alignmentcited by 4
  10. 2026AscendOptimizer: Episodic Agent for Ascend NPU Operator Optimizationcited by 3
  11. 2026SOLE-R1: Video-Language Reasoning as the Sole Reward for On-Robot Reinforcement Learningcited by 3
  12. 2026RDA: Reward Design Agent for Reinforcement Learningcited by 3
  13. 2026RARM: Confidence-Gated Progress Reward Modeling for RL in Manipulationcited by 3
  14. 2025AURA: Autonomous Upskilling with Retrieval-Augmented Agentscited by 2
  15. 2025Mash, Spread, Slice! Learning to Manipulate Object States via Visual Spatial Progresscited by 2
  16. 2026Rewarding DINO: Predicting Dense Rewards with Vision Foundation Modelscited by 2
  17. 2026SARM2: Multi-Task Stage Aware Reward Modeling for Self Improving Robotic Manipulationcited by 2
  18. 2026InSight: Self-Guided Skill Acquisition via Steerable VLAscited by 2
  19. 2026DenseReward: Dense Reward Learning via Failure Synthesis for Robotic Manipulationcited by 2
  20. 2025VITA: Zero-Shot Value Functions via Test-Time Adaptation of Vision-Language Modelscited by 1
  21. 2025When a Robot is More Capable than a Human: Learning from Constrained Demonstratorscited by 1
  22. 2026MVR: Multi-view Video Reward Shaping for Reinforcement Learningcited by 1
  23. 2026Beyond Pixels: Learning Invariant Rewards for Real-World Robotics From a Few Demonstrationscited by 1
  24. 2026Learning Process Rewards via Success Visitation Matching for Efficient RLcited by 1
  25. 2026Stage-Transition Dense Reward Modeling for Reinforcement Learningcited by 1
  26. 2026CounterAlign: Counterfactual Supervision for Vision-Language-Action Modelscited by 1
  27. 2026SUN: Agentic Robot Policy Learning with Persistent Task Programscited by 1
  28. 2026Reward-Zero: Language Embedding Driven Implicit Reward Mechanisms for Reinforcement Learningcited by 0
  29. 2026HandelBot: Real-World Piano Playing via Fast Adaptation of Dexterous Robot Policiescited by 0
  30. 2026AgenticRL: Agentic Reinforcement Learning with Self-Refinement for Complex UAV Navigationcited by 0
  31. 2026DexPIE: Stable Dexterous Policy Improvement from Real-World Experiencecited by 0
  32. 2026Language-Critique Imitation Learning from Suboptimal Demonstrationscited by 0
  33. 2026SynCity 3000: Bootstrapping Scene-Scale 3D Diffusioncited by 0
  34. 2026Optimal Transport Q-Learning for Flow Policy Steering and Accelerationcited by 0
  35. 2026HELP: Human-Efficient Large-Scale Robot Post-Training with Rollout Segmentationcited by 0
  36. 2026Same Trajectory, Contradictory Rewards (ROBORMBENCH): Paraphrase Fragility in Vision Language Reward Modelscited by 0
  37. 2026Finishing the Task Is Not Enough: Evaluating Agent Resilience and Considerate Participation under Accumulating Challengecited by 0
  38. 2026SynthDemo-RL: Breaking the Zero-Reward Barrier in VLA Adaptation with LLM-Guided Synthetic Demonstrationscited by 0

50 most cited of 60 All 60 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024Video-Language Critic: Transferable Reward Functions for Language-Conditioned Robotics
  2. 2025Masked IRL: LLM-Guided Reward Disambiguation from Demonstrations and Language
  3. 2019From Language to Goals: Inverse Reinforcement Learning for Vision-Based Instruction Following
  4. 2026From Demonstrations to Rewards: Test-Time Prompt Optimization for VLM Reward Models
  5. 2026Large Reward Models: Generalizable Online Robot Reward Generation with Vision-Language Models
  6. 2024Learning Manipulation Skills through Robot Chain-of-Thought with Sparse Failure Guidance
  7. 2026From Pretraining to Proficiency: Real-World Subtask RL for Long-Horizon Manipulation with Minimal Human Intervention
  8. 2023Text2Reward: Reward Shaping with Language Models for Reinforcement Learning