mlsanityData to 7 Oct 2026

Learning Task Decomposition with Ordered Memory Policy Network

arXiv 2103.10972 ↗First posted 19 Mar 2021cs.LG cs.AIcited by 15

Not on the map. Why?

Abstract (arXiv)

Many complex real-world tasks are composed of several levels of sub-tasks. Humans leverage these hierarchical structures to accelerate the learning process and achieve better generalization. In this work, we study the inductive bias and propose Ordered Memory Policy Network (OMPN) to discover subtask hierarchy by learning from demonstration. The discovered subtask hierarchy could be used to perform task decomposition, recovering the subtask boundaries in an unstruc-tured demonstration. Experiments on Craft and Dial demonstrate that our modelcan achieve higher task decomposition performance under both unsupervised and weakly supervised settings, comparing with strong baselines. OMPN can also bedirectly applied to partially observable environments and still achieve higher task decomposition performance. Our visualization further confirms that the subtask hierarchy can emerge in our model.

Key facts. Cited by 15 papers in the map (data through 7 Oct 2026). Builds on 16 of its 28 references that are in the map. First posted on arXiv 19 Mar 2021.

Builds on

16 of 28 references.

References · 16

  1. 2018Data-Efficient Hierarchical Reinforcement Learningcited by 436
  2. 2016The Option-Critic Architecturecited by 433
  3. 2019Relay Policy Learning: Solving Long-Horizon Tasks via Imitation and Reinforcement Learningcited by 380
  4. 2017FeUdal Networks for Hierarchical Reinforcement Learningcited by 328
  5. 2019Learning Latent Plans from Playcited by 310
  6. 2016Modular Multitask Reinforcement Learning with Policy Sketchescited by 200
  7. 2017DART: Noise Injection for Robust Imitation Learningcited by 187
  8. 2018Variational Option Discovery Algorithmscited by 121
  9. 2016Hierarchical Multiscale Recurrent Neural Networkscited by 102
  10. 2018Ordered Neurons: Integrating Tree Structures into Recurrent Neural Networkscited by 101
  11. 2014A Clockwork RNNcited by 70
  12. 2018TACO: Learning Task Decomposition via Temporal Alignment for Controlcited by 61
4 more
  1. 2017Multi-Level Discovery of Deep Optionscited by 50
  2. 2017DDCO: Discovery of Deep Continuous Options for Robot Learning from Demonstrationscited by 43
  3. 2020Learning to Combine Top-Down and Bottom-Up Signals in Recurrent Neural Networks with Attention over Modulescited by 38
  4. 2019Ordered Memorycited by 20

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 2; 2022: 6; 2023: 5; 2024: 0; 2025: 0; 2026: 2.

15 papers.

Sort:

By area:

  1. 2023Multi-Stage Cable Routing through Hierarchical Imitation Learningcited by 59
  2. 2022Skill-based Model-based Reinforcement Learningcited by 50
  3. 2021Adversarial Skill Chaining for Long-Horizon Robot Manipulation via Terminal State Regularizationcited by 44
  4. 2023Sequential Dexterity: Chaining Dexterous Policies for Long-Horizon Manipulationcited by 43
  5. 2022Generalizing Goal-Conditioned Reinforcement Learning with Variational Causal Reasoningcited by 33
  6. 2023Seeing is not Believing: Robust Reinforcement Learning against Spurious Correlationcited by 25
  7. 2022Learning Options via Compressioncited by 22
  8. 2022Unsupervised Learning of Temporal Abstractions with Slot-based Transformerscited by 14
  9. 2022Toward Modeling Creative Processes for Algorithmic Paintingcited by 9
  10. 2023Learning Rational Subgoals from Demonstrations and Instructionscited by 5
  11. 2022Possibility Before Utility: Learning And Using Hierarchical Affordancescited by 4
  12. 2026Learning Abstractions for Hierarchical Planning in Program-Synthesis Agentscited by 3
3 more
  1. 2021A Simple Approach to Continual Learning by Transferring Skill Parameterscited by 2
  2. 2023Boosting Reinforcement Learning and Planning with Demonstrations: A Surveycited by 1
  3. 2026Unsupervised Hierarchical Skill Discoverycited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Identifying Selections for Unsupervised Subtask Discovery
  2. 2025Conditional Diffusion Model for Multi-Agent Dynamic Task Decomposition
  3. 2024Task-Oriented Hierarchical Object Decomposition for Visuomotor Control
  4. 2025Internal Chain-of-Thought: Empirical Evidence for Layer-wise Subtask Scheduling in LLMs
  5. 2022Generalization with Lossy Affordances: Leveraging Broad Offline Data for Learning Visuomotor Tasks
  6. 2026On-Policy Attention Linearization
  7. 2026Reinforcement Learning with Decomposed Subtasks
  8. 2022Active Predictive Coding: A Unified Neural Framework for Learning Hierarchical World Models for Perception and Planning