mlsanityData to 7 Oct 2026

Sample-Efficient Behavior Cloning Using General Domain Knowledge

arXiv 2501.16546 ↗First posted 27 Jan 2025cs.AIcited by 2

On the map: Offline RL and imitation learning from data → Imitation learning methods

Abstract (arXiv)

Behavior cloning has shown success in many sequential decision-making tasks by learning from expert demonstrations, yet they can be very sample inefficient and fail to generalize to unseen scenarios. One approach to these problems is to introduce general domain knowledge, such that the policy can focus on the essential features and may generalize to unseen states by applying that knowledge. Although this knowledge is easy to acquire from the experts, it is hard to be combined with learning from individual examples due to the lack of semantic structure in neural networks and the time-consuming nature of feature engineering. To enable learning from both general knowledge and specific demonstration trajectories, we use a large language model's coding capability to instantiate a policy structure based on expert domain knowledge expressed in natural language and tune the parameters in the policy with demonstrations. We name this approach the Knowledge Informed Model (KIM) as the structure reflects the semantics of expert knowledge. In our experiments with lunar lander and car racing tasks, our approach learns to solve the tasks with as few as 5 demonstrations and is robust to action noise, outperforming the baseline model without domain knowledge. This indicates that with the help of large language models, we can incorporate domain knowledge into the structure of the policy, increasing sample efficiency for behavior cloning.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Placed in Imitation learning methods, one of 13 directions in Offline RL and imitation learning from data. Builds on 33 of its 38 references that are in the map. First posted on arXiv 27 Jan 2025.

Builds on

33 of 38 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024Efficient Imitation Learning with Conservative World Models21 May
  2. 2018An Algorithmic Perspective on Imitation Learning16 Nov

Other references · 31

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  3. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  4. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  5. 2023LLM+P: Empowering Large Language Models with Optimal Planning Proficiencycited by 496
  6. 2022Git Re-Basin: Merging Models modulo Permutation Symmetriescited by 460
  7. 2024Mobile ALOHA: Learning Bimanual Mobile Manipulation with Low-Cost Whole-Body Teleoperationcited by 377
  8. 2024Alignment faking in large language modelscited by 316
  9. 2023Parting with Misconceptions about Learning-based Vehicle Motion Planningcited by 238
  10. 2022Structure-based Drug Design with Equivariant Diffusion Modelscited by 213
  11. 2022Model-Based Imitation Learning for Urban Drivingcited by 149
  12. 2024RL-VLM-F: Reinforcement Learning from Vision Language Foundation Model Feedbackcited by 147
19 more
  1. 2024ALOHA Unleashed: A Simple Recipe for Robot Dexteritycited by 128
  2. 2023An Expert's Guide to Training Physics-informed Neural Networkscited by 119
  3. 2024WorldCoder, a Model-Based LLM Agent: Building World Models by Writing Code and Interacting with the Environmentcited by 99
  4. 2023Decomposing the Generalization Gap in Imitation Learning for Visual Robotic Manipulationcited by 78
  5. 2023LOTUS: Continual Imitation Learning for Robot Manipulation Through Unsupervised Skill Discoverycited by 63
  6. 2023NeRF in the Palm of Your Hand: Corrective Augmentation for Robotics via Novel-View Synthesiscited by 56
  7. 2024Code Repair with LLMs gives an Exploration-Exploitation Tradeoffcited by 50
  8. 2021People construct simplified mental representations to plancited by 45
  9. 2024JUICER: Data-Efficient Imitation Learning for Robotic Assemblycited by 36
  10. 2023Large Language Model as a Policy Teacher for Training Reinforcement Learning Agentscited by 30
  11. 2024Deep SE(3)-Equivariant Geometric Reasoning for Precise Placement Taskscited by 19
  12. 2023Look Beneath the Surface: Exploiting Fundamental Symmetry for Sample-Efficient Offline RLcited by 18
  13. 2019Feature Selection and Feature Extraction in Pattern Analysis: A Literature Reviewcited by 15
  14. 2024Learning with Language-Guided State Abstractionscited by 11
  15. 2022Visual Backtracking Teleoperation: A Data Collection Protocol for Offline Image-Based Reinforcement Learningcited by 10
  16. 2024Bootstrapping Cognitive Agents with a Large Language Modelcited by 9
  17. 2022Robust Imitation of a Few Demonstrations with a Backwards Modelcited by 7
  18. 2023PDSketch: Integrated Planning Domain Programming and Learningcited by 7
  19. 2024Data Efficient Behavior Cloning for Fine Manipulation via Continuity-based Corrective Labelscited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026InterPReT: Interactive Policy Restructuring and Training Enable Effective Imitation Learning from Laypersonscited by 2
  2. 2026Iterative Policy Refinement through Semantic Rollout Analysiscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Is Behavior Cloning All You Need? Understanding Horizon in Imitation Learning
  2. 2021Implicit Behavioral Cloning
  3. 2025Posterior Behavioral Cloning: Pretraining BC Policies for Efficient RL Finetuning
  4. 2024Imitation Learning from Suboptimal Demonstrations via Meta-Learning An Action Ranker
  5. 2025Blindfolded Experts Generalize Better: Insights from Robotic Manipulation and Videogames
  6. 2020Imitating Unknown Policies via Exploration
  7. 2020Learning to Generalize Across Long-Horizon Tasks from Human Demonstrations
  8. 2024Stable-BC: Controlling Covariate Shift with Stable Behavior Cloning