mlsanityData to 14 Oct 2026

Neuro-Symbolic Imitation Learning: Discovering Symbolic Abstractions for Skill Learning

arXiv 2503.21406 ↗First posted 27 Mar 2025cs.AI cs.LG cs.ROcited by 4

On the map: Offline RL and imitation learning from data → Robot manipulation imitation learning

Abstract (arXiv)

Imitation learning is a popular method for teaching robots new behaviors. However, most existing methods focus on teaching short, isolated skills rather than long, multi-step tasks. To bridge this gap, imitation learning algorithms must not only learn individual skills but also an abstract understanding of how to sequence these skills to perform extended tasks effectively. This paper addresses this challenge by proposing a neuro-symbolic imitation learning framework. Using task demonstrations, the system first learns a symbolic representation that abstracts the low-level state-action space. The learned representation decomposes a task into easier subtasks and allows the system to leverage symbolic planning to generate abstract plans. Subsequently, the system utilizes this task decomposition to learn a set of neural skills capable of refining abstract plans into actionable robot commands. Experimental results in three simulated robotic environments demonstrate that, compared to baselines, our neuro-symbolic approach increases data efficiency, improves generalization capabilities, and facilitates interpretability.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Placed in Robot manipulation imitation learning, one of 13 directions in Offline RL and imitation learning from data. Builds on 33 of its 68 references that are in the map. First posted on arXiv 27 Mar 2025.

Builds on

33 of 68 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023Human-in-the-Loop Task and Motion Planning for Imitation Learning24 Oct

Other references · 32

  1. 2020robosuite: A Modular Simulation Framework and Benchmark for Robot Learningcited by 511
  2. 2018An Algorithmic Perspective on Imitation Learningcited by 459
  3. 2020Integrated Task and Motion Planningcited by 290
  4. 2023A Survey of Imitation Learning: Algorithms, Recent Developments, and Challengescited by 148
  5. 2018SDRL: Interpretable and Data-efficient Deep Reinforcement Learning Leveraging Symbolic Planningcited by 78
  6. 2018TACO: Learning Task Decomposition via Temporal Alignment for Controlcited by 62
  7. 2017Classical Planning in Deep Latent Space: Bridging the Subsymbolic-Symbolic Boundarycited by 58
  8. 2021Imitation Learning: Progress, Taxonomies and Challengescited by 58
  9. 2022Learning Neuro-Symbolic Skills for Bilevel Planningcited by 58
  10. 2018PEORL: Integrating Symbolic Planning and Hierarchical Reinforcement Learning for Robust Decision-Makingcited by 52
  11. 2022Predicate Invention for Bilevel Planningcited by 49
  12. 2021Learning Neuro-Symbolic Relational Transition Models for Bilevel Planningcited by 43
20 more
  1. 2019Learning First-Order Symbolic Representations for Planning from the Structure of the State Spacecited by 38
  2. 2022LEAGUE: Guided Skill Learning and Abstraction for Long-Horizon Manipulationcited by 34
  3. 2019Unsupervised Grounding of Plannable First-Order Logic Representation from Imagescited by 24
  4. 2023Neurosymbolic Reinforcement Learning and Planning: A Surveycited by 24
  5. 2021Learning First-Order Representations for Planning from Black-Box States: New Resultscited by 23
  6. 2021SKID RAW: Skill Discovery from Raw Trajectoriescited by 22
  7. 2021Classical Planning in Deep Latent Spacecited by 22
  8. 2022Leveraging Approximate Symbolic Models for Reinforcement Learning via Skill Diversitycited by 22
  9. 2021Discovering State and Action Abstractions for Generalized Task and Motion Planningcited by 18
  10. 2021Guided Imitation of Task and Motion Planningcited by 18
  11. 2017Efficient, Safe, and Probably Approximately Complete Learning of Action Modelscited by 16
  12. 2019Learning Portable Representations for High-Level Planningcited by 11
  13. 2020Planning from Pixels in Atari with Learned Symbolic Representationscited by 11
  14. 2021Discovering User-Interpretable Capabilities of Black-Box Planning Agentscited by 10
  15. 2017Active Exploration for Learning Symbolic Representationscited by 8
  16. 2021Evolutionary Training and Abstraction Yields Algorithmic Generalization of Neural Computerscited by 5
  17. 2023Discovering Predictive Relational Object Symbols with Symbolic Attentive Layerscited by 3
  18. 2023Learning Type-Generalized Actions for Symbolic Planningcited by 2
  19. 2021Creativity of AI: Hierarchical Planning Model Learning for Facilitating Deep Reinforcement Learningcited by 1
  20. 2024Efficient Symbolic Planning with Viewscited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 2.

4 papers.

  1. 2025Few-Shot Neuro-Symbolic Imitation Learning for Long-Horizon Planning and Actingcited by 2
  2. 2025Learning Hierarchical Domain Models Through Environment-Grounded Interactioncited by 1
  3. 2026Build on Priors: Vision--Language--Guided Neuro-Symbolic Imitation Learning for Data-Efficient Real-World Robot Manipulationcited by 0
  4. 2026Regression Test Selection for Updated Capability Modules in Compositional ML Systems via Atomic-Quality Probescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2022Learning and Retrieval from Prior Data for Skill-based Imitation Learning
  2. 2024Neuro-Symbolic Skill Discovery for Conditional Multi-Level Planning
  3. 2024VisualPredicator: Learning Abstract World Models with Neuro-Symbolic Predicates for Robot Planning
  4. 2024Robotic Imitation of Human Actions
  5. 2025Symskill: Symbol and Skill Co-Invention for Data-Efficient and Reactive Long-Horizon Manipulation
  6. 2026Learning Compositional Symbolic Task Rules from Demonstrations with Inductive Logic Programming
  7. 2026From Local Corrections to Generalized Skills: Improving Neuro-Symbolic Policies with MEMO
  8. 2019Continuous Relaxation of Symbolic Planner for One-Shot Imitation Learning