Overview

Policy distillation studies how to transfer useful behavior from a teacher policy to a student that may be smaller, less informed, or easier to deploy. Common approaches match teacher actions or probability distributions using supervised learning, KL objectives, on-policy rollouts, offline data, or reinforcement-learning feedback; multi-teacher mixtures, generative students, and interpretable models extend this to diverse skills and compact controllers. The methods target lower inference costs, specialist-skill integration, and transfer from privileged state to robot sensors or smaller language models, with evaluation in settings such as offline control, multi-agent tasks, reasoning benchmarks, and real-robot manipulation and navigation. A central tension is that imitation can copy flawed or inaccessible teacher behavior and lose valid action modes, so students need guidance that matches their deployment information while retaining enough exploration and feedback to generalize or improve on the teacher.

Latest papers

CardsList

56 papers11 papers in the last four weeks, down 8% on the four weeks before. 0.1% of all new papers.

  1. Discovering Programmatic Policies from Reinforcement Learning-Based Traffic Signal Controllers

    Nov 15, 2026Lindong Xie, Yang Zhang, Beiyu Song +3Traffic Signal ControlPolicy Distillation

  2. Program Distillation: Transferring LLM Capability from System 2 to System 0

    Nov 14, 2026Pujun Zheng, Wanying Ren, Guoxiu HePolicy DistillationKnowledge Distillation

  3. ReShap: Interpretable Reinforcement Learning for High-Dimensional Multi-Action Settings

    Nov 11, 2026Yiyu Qian, Pengfei Guo, Liyuan ZhaoInterpretability in RLReinforcement Learning

  4. Value-oriented policy distillation for interpretable wind farm power control

    Nov 5, 2026Max Bruninx, Timothy Verstraeten, Jan Decuyper +2Power SystemsPolicy Distillation

  5. Using Transformers for Teaching and Learning via Symbolic Distillation of Neural A-life Agents

    Nov 5, 2026Shuwei Wang, Vadim Bulitko, Matthew R. G. Brown +5Policy DistillationArtificial Life

  6. Gaussian Distillation of Flow Policies

    Oct 29, 2026Youssef Saied, Ziyad Sheebaelhamd, Claire Vernade +1Robot Policy LearningFlow-Based Policy Learning

  7. TerraTransfer++: Demonstration-Free End-to-End Planner

    Oct 21, 2026Zikang Xiong, Weixin Li, Zhouchonghao Wu +6End-to-End Autonomous DrivingAutonomous Driving Planning

  8. Value-Rectified Distillation for Flow-based Offline Reinforcement Learning

    Oct 21, 2026Ke Jiang, Wen Jiang, Yoshinobu Kawahara +1Offline RLPolicy Distillation

  9. ReDOA: Real-World Distillation via Observation Alignment for Image-Based Manipulation Policy Learning

    Oct 19, 2026Kotaro Kajitsuka, Masaki Murooka, Fumio KanehiroVisuomotor Policy LearningRobotic Manipulation

  10. Harness Policy Distillation: Aligning Agent Policies with Runtime Semantics

    Oct 14, 2026Junze Zhu, Weihao Chen, Zhiyang Zhou +2LLM Agent TrainingPolicy Distillation

  11. TAME-Nav: Temporal Asymmetric Mixture-of-Experts Distillation for Agile and Versatile Robot Navigation

    Oct 9, 2026Lingjie Zhang, Jun Xue, Shanze Wang +4Robot NavigationRobot Policy Learning

  12. Composing What Each Teacher Learned: Delta Distillation for Multi-Teacher On-Policy Learning

    Oct 7, 2026Yuanda Xu, Hejian Sang, Zhengze Zhou +4Multi-Teacher Knowledge DistillationOn-Policy Distillation

  13. Teaching a Robot Dog New Tricks: Diverse Quadruped Skills via Combined Reinforcement and Imitation Learning with Adversarial Task Selection

    Oct 6, 2026Lemon Foxmere, Anthony Furman, Yizheng Du +3Robot Skill LearningMulti-Task RL

  14. iGPC: Generative Motion Priors for Object-Aware Humanoid Interaction

    Oct 6, 2026Anujith Muraleedharan, Abdul Ahad Butt, Nolan Fey +5Contact-Rich Robotic ManipulationHumanoid Robot Control

  15. Interpolated Policy Distillation: A Controllable Continuum Between Off-Policy and On-Policy Distillation

    Sep 29, 2026Youxu Shi, Yifan Sun, Dacheng Yin +5Language Model DistillationOn-Policy Distillation

  16. Distilling Privileged Control Barrier Functions into RGB-Only Safety Filters for Dynamic Visual Navigation

    Sep 29, 2026Seungyeon Yoo, Gawon Lee, Seungwoo Jung +2Control Barrier FunctionsRobotics

  17. Inductive Feedback for Mixed-Policy Distillation

    Sep 28, 2026Amir Moeini, Huaijiang Zhu, Daniel Havir +1Language Model DistillationOn-Policy Distillation

  18. Beyond Token Alignment: Event Completion for Cross-Tokenizer On-Policy Distillation

    Sep 28, 2026Jiacheng Liu, Jingwei Song, Qituan Zhang +2Cross-Tokenizer Knowledge DistillationOn-Policy Distillation

  19. Experience Funnel: A State-Policy Alternating Loop for Self-Evolving Agents

    Sep 8, 2026Wenbo Gao, Zhaomou Song, Zhiyuan Ji +7Continual Learning for LLM AgentsAgent Skill Learning

  20. Selective Agent Guidance via Entropy: Learning Autonomous Policies from Imperfect VLM Teachers

    Sep 1, 2026Giovanni Bonetta, Matteo Merler, Davide Zago +2Uncertainty SamplingVisual RL

  21. SOLO: Stable Omni-terrain Long-Horizon Perceptive Humanoid Locomotion

    Aug 27, 2026Pihai Sun, Gang Han, Jingkai Sun +17Humanoid Robot Locomotion3D Reconstruction

  22. EDGE: Experience-Distillation for Guided Exploration in Agentic Reinforcement Learning

    Aug 22, 2026Can Xie, Yuyi Zhou, Wen Yang +5Agentic RLRL Exploration

  23. CIDER: Continual Interactive Distillation for Embodied Reinforcement Learning

    Aug 22, 2026Houlin Li, Minghui Xu, Guo Xu +8Continual Robot LearningContinual RL

  24. Adaptive Mixing of Policies from Searching and Policies from Learning

    Aug 16, 2026Gavin B. RensReinforcement LearningPolicy Learning