mlsanityData to 14 Oct 2026

Learning Top-k Subtask Planning Tree based on Discriminative Representation Pre-training for Decision Making

arXiv 2312.11027 ↗First posted 18 Dec 2023cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Many complicated real-world tasks can be broken down into smaller, more manageable parts, and planning with prior knowledge extracted from these simplified pieces is crucial for humans to make accurate decisions. However, replicating this process remains a challenge for AI agents and naturally raises two questions: How to extract discriminative knowledge representation from priors? How to develop a rational plan to decompose complex problems? Most existing representation learning methods employing a single encoder structure are fragile and sensitive to complex and diverse dynamics. To address this issue, we introduce a multiple-encoder and individual-predictor regime to learn task-essential representations from sufficient data for simple subtasks. Multiple encoders can extract adequate task-relevant dynamics without confusion, and the shared predictor can discriminate the task characteristics. We also use the attention mechanism to generate a top-k subtask planning tree, which customizes subtask execution plans in guiding complex decisions on unseen tasks. This process enables forward-looking and globality by flexibly adjusting the depth and width of the planning tree. Empirical results on a challenging platform composed of some basic simple tasks and combinatorially rich synthetic tasks consistently outperform some competitive baselines and demonstrate the benefits of our design.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 41 of its 65 references that are in the map. First posted on arXiv 18 Dec 2023.

Builds on

41 of 65 references.

References · 41

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  3. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,548
  4. 2023Voyager: An Open-Ended Embodied Agent with Large Language Modelscited by 2,191
  5. 2018Soft Actor-Critic Algorithms and Applicationscited by 1,240
  6. 2022ProgPrompt: Generating Situated Robot Task Plans using Large Language Modelscited by 611
  7. 2022BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learningcited by 523
  8. 2016Target-driven Visual Navigation in Indoor Scenes using Deep Reinforcement Learningcited by 366
  9. 2019On the Utility of Learning about Humans for Human-AI Coordinationcited by 349
  10. 2020Integrated Task and Motion Planningcited by 290
  11. 2021Multi-Task Reinforcement Learning with Context-based Representationscited by 174
  12. 2019Planning with Goal-Conditioned Policiescited by 149
29 more
  1. 2020Multi-Task Reinforcement Learning with Soft Modularizationcited by 143
  2. 2021Goal-Conditioned Reinforcement Learning with Imagined Subgoalscited by 134
  3. 2020World Model as a Graph: Learning Latent Landmarks for Planningcited by 69
  4. 2020Long-Horizon Visual Planning with Goal-Conditioned Hierarchical Predictorscited by 62
  5. 2022Learning Neuro-Symbolic Skills for Bilevel Planningcited by 58
  6. 2023TPTU: Large Language Model-based AI Agents for Task Planning and Tool Usagecited by 55
  7. 2018Learning Actionable Representations with Goal-Conditioned Policiescited by 54
  8. 2018Hierarchical Reinforcement Learning for Zero-shot Generalization with Subtask Dependenciescited by 43
  9. 2022Robust Task Representations for Offline Meta-Reinforcement Learning via Contrastive Learningcited by 35
  10. 2022A Survey on Recent Advances and Challenges in Reinforcement Learning Methods for Task-Oriented Dialogue Policy Learningcited by 30
  11. 2020Sub-Goal Trees -- a Framework for Goal-Based Reinforcement Learningcited by 29
  12. 2021Evaluating the Robustness of Collaborative Agentscited by 25
  13. 2021CARL: A Benchmark for Contextual and Adaptive Reinforcement Learningcited by 21
  14. 2021Search-Based Task Planning with Learned Skill Effect Models for Lifelong Robotic Manipulationcited by 20
  15. 2020Transferable Task Execution from Pixels through Deep Planning Domain Learningcited by 17
  16. 2022Reward-Respecting Subtasks for Model-Based Reinforcement Learningcited by 16
  17. 2022Model-Based Offline Meta-Reinforcement Learning with Regularizationcited by 15
  18. 2022LDSA: Learning Dynamic Subtask Assignment in Cooperative Multi-Agent Reinforcement Learningcited by 15
  19. 2022Discrete Factorial Representations as an Abstraction for Goal Conditioned Reinforcement Learningcited by 11
  20. 2021Discovering User-Interpretable Capabilities of Black-Box Planning Agentscited by 10
  21. 2021PantheonRL: A MARL Library for Dynamic Training Interactionscited by 10
  22. 2023Planning with Sequence Models through Iterative Energy Minimizationcited by 8
  23. 2022Contextual Transformer for Offline Meta Reinforcement Learningcited by 7
  24. 2022Learning Task Embeddings for Teamwork Adaptation in Multi-Agent Reinforcement Learningcited by 5
  25. 2022Intention-Aware Navigation in Crowds with Extended-Space POMDP Planningcited by 4
  26. 2023On Context Distribution Shift in Task Representation Learning for Offline Meta RLcited by 4
  27. 2020REMAX: Relational Representation for Multi-Agent Explorationcited by 3
  28. 2022Discovering Generalizable Spatial Goal Representations via Graph-based Active Reward Learningcited by 3
  29. 2023Learning to Collaborate by Grouping: a Consensus-oriented Strategy for Multi-agent Reinforcement Learningcited by 1

Built on by

Not cited yet.