mlsanityData to 7 Oct 2026

RDA: Reward Design Agent for Reinforcement Learning

arXiv 2606.01672 ↗First posted 1 Jun 2026cs.LGcited by 1

On the map: LLM agents and tool use → Reinforcement learning for LLM agents

Abstract (arXiv)

Reinforcement learning has enabled the acquisition of impressive robotic skills, but typically requires hand-crafted reward functions that are slow to design and difficult to align with human intentions. Recent work, such as Eureka, automates reward design by using an LLM to iteratively generate and refine reward code from task descriptions. However, they rely on coarse feedback signals such as success rate, which provide little semantic insight into the learned behavior. As a result, their trained policies achieve the final goal but are frequently poorly aligned with task instructions. We introduce the Reward Design Agent (RDA), a VLM-based agentic framework that injects semantic understanding into reward design. RDA decomposes tasks, visually evaluates trajectories, summarizes failure modes, and iteratively revises reward code to better align with task instructions. Across 12 tabletop manipulation tasks from ManiSkill and 4 whole-body manipulation tasks from HumanoidBench, RDA produces policies substantially more instruction-aligned than those of other baselines, while achieving comparable task success rates. Videos and the generated reward code are available on https://nitinkamra1992.github.io/reward-design-agent.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use. Builds on 44 of its 52 references that are in the map. First posted on arXiv 1 Jun 2026.

Builds on

44 of 52 references.

References · 44

  1. 2017Deep reinforcement learning from human preferencescited by 3,846
  2. 2016Generative Adversarial Imitation Learningcited by 1,576
  3. 2018Soft Actor-Critic Algorithms and Applicationscited by 1,179
  4. 2024The AI Scientist: Towards Fully Automated Open-Ended Scientific Discoverycited by 1,002
  5. 2021Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learningcited by 882
  6. 2023Eureka: Human-Level Reward Design via Coding Large Language Modelscited by 542
  7. 2019Learning agile and dynamic motor skills for legged robotscited by 402
  8. 2022VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Trainingcited by 384
  9. 2023Language to Rewards for Robotic Skill Synthesiscited by 307
  10. 2021PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-trainingcited by 298
  11. 2023Reward Design with Language Modelscited by 257
  12. 2024ManiSkill3: GPU Parallelized Robotics Simulation and Rendering for Generalizable Embodied AIcited by 216
32 more
  1. 2018Discriminator-Actor-Critic: Addressing Sample Inefficiency and Reward Bias in Adversarial Imitation Learningcited by 203
  2. 2023Vision-Language Models are Zero-Shot Reward Models for Reinforcement Learningcited by 155
  3. 2024Deep Reinforcement Learning for Robotics: A Survey of Real-World Successescited by 153
  4. 2024RL-VLM-F: Reinforcement Learning from Vision Language Foundation Model Feedbackcited by 147
  5. 2023RoboCLIP: One Demonstration is Enough to Learn Robot Policiescited by 128
  6. 2018Variational Discriminator Bottleneck: Improving Imitation Learning, Inverse RL, and GANs by Constraining Information Flowcited by 103
  7. 2021Reward (Mis)design for Autonomous Drivingcited by 102
  8. 2024HumanoidBench: Simulated Humanoid Benchmark for Whole-Body Locomotion and Manipulationcited by 101
  9. 2024MaskedMimic: Unified Physics-Based Character Control Through Masked Motion Inpaintingcited by 101
  10. 2022Few-Shot Preference Learning for Human-in-the-Loop RLcited by 85
  11. 2024Vision Language Models are In-Context Value Learnerscited by 84
  12. 2024DrEureka: Language Model Guided Sim-To-Real Transfercited by 78
  13. 2023Video Prediction Models as Rewards for Reinforcement Learningcited by 76
  14. 2018Learning Navigation Behaviors End-to-End with AutoRLcited by 70
  15. 2024SimBa: Simplicity Bias for Scaling Up Parameters in Deep Reinforcement Learningcited by 70
  16. 2025A Vision-Language-Action-Critic Model for Robotic Real-World Reinforcement Learningcited by 66
  17. 2023Vision-Language Models as a Source of Rewardscited by 65
  18. 2023Language Instructed Reinforcement Learning for Human-AI Coordinationcited by 61
  19. 2025ReWiND: Language-Guided Rewards Teach Robot Policies without New Demonstrationscited by 52
  20. 2026RoboReward: General-Purpose Vision-Language Reward Models for Roboticscited by 46
  21. 2021Learning Multimodal Rewards from Rankingscited by 41
  22. 2019Evolving Rewards to Automate Reinforcement Learningcited by 28
  23. 2021Learning Reward Functions from Scale Feedbackcited by 24
  24. 2024Trajectory Improvement and Reward Learning from Comparative Language Feedbackcited by 17
  25. 2025Enhancing Rating-Based Reinforcement Learning to Effectively Leverage Feedback from Large Vision-Language Modelscited by 15
  26. 2024VICtoR: Learning Hierarchical Vision-Instruction Correlation Rewards for Long-horizon Manipulationcited by 11
  27. 2025Towards Improving Reward Design in RL: A Reward Alignment Metric for RL Practitionerscited by 10
  28. 2025GROVE: A Generalized Reward for Learning Open-Vocabulary Physical Skillcited by 9
  29. 2026FlashSAC: Fast and Stable Off-Policy Reinforcement Learning for High-Dimensional Robot Controlcited by 5
  30. 2025Policy Learning from Large Vision-Language Model Feedback without Reward Modelingcited by 4
  31. 2025Automated Reward Design for Gran Turismocited by 2
  32. 2026LocoVLM: Grounding Vision and Language for Adapting Versatile Legged Locomotion Policiescited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026InterEvolve: Test-Time Evolution of Reward Programs for Humanoid Loco-Manipulationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Large Reward Models: Generalizable Online Robot Reward Generation with Vision-Language Models
  2. 2026Hierarchical Reward Design from Language: Enhancing Alignment of Agent Behavior with Human Specifications
  3. 2026MimicAgent: Quadruped Skills via Prompt-to-Trajectory Generation
  4. 2025Training-free Generation of Temporally Consistent Rewards from VLMs
  5. 2025Robo-Dopamine: General Process Reward Modeling for High-Precision Robotic Manipulation
  6. 2025Reward Evolution with Graph-of-Thoughts: A Bi-Level Language Model Framework for Reinforcement Learning
  7. 2026Exploring Reasoning Reward Model for Agents
  8. 2026SARM2: Multi-Task Stage Aware Reward Modeling for Self Improving Robotic Manipulation