Fast Inference and Transfer of Compositional Task Structures for Few-shot Task Generalization
Not on the map. Why?
Abstract (arXiv)
We tackle real-world problems with complex structures beyond the pixel-based game or simulator. We formulate it as a few-shot reinforcement learning problem where a task is characterized by a subtask graph that defines a set of subtasks and their dependencies that are unknown to the agent. Different from the previous meta-rl methods trying to directly infer the unstructured task embedding, our multi-task subtask graph inferencer (MTSGI) first infers the common high-level task structure in terms of the subtask graph from the training tasks, and use it as a prior to improve the task inference in testing. Our experiment results on 2D grid-world and complex web navigation domains show that the proposed method can learn and leverage the common underlying structure of the tasks for faster adaptation to the unseen tasks than various existing algorithms such as meta reinforcement learning, hierarchical reinforcement learning, and other heuristic agents.
Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 23 of its 46 references that are in the map. First posted on arXiv 25 May 2022.
Builds on
23 of 46 references.
References · 23
- 2017Proximal Policy Optimization Algorithmscited by 15,215
- 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
- 2018On First-Order Meta-Learning Algorithmscited by 1,018
- 2017GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networkscited by 834
- 2016RL2: Fast Reinforcement Learning via Slow Reinforcement Learningcited by 503
- 2016Learning to reinforcement learncited by 377
- 2018Reinforcement Learning on Web Interfaces Using Workflow-Guided Explorationcited by 284
- 2018Probabilistic Model-Agnostic Meta-Learningcited by 256
- 2016Modular Multitask Reinforcement Learning with Policy Sketchescited by 203
- 2018Bayesian Model-Agnostic Meta-Learningcited by 194
- 2020Acme: A Research Framework for Distributed Reinforcement Learningcited by 146
- 2018Meta-Reinforcement Learning of Structured Exploration Strategiescited by 130
11 more
- 2017Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learningcited by 98
- 2017Neural Task Programming: Learning to Generalize Across Hierarchical Taskscited by 76
- 2017Gated-Attention Architectures for Task-Oriented Language Groundingcited by 72
- 2018Neural Task Graphs: Generalizing to Unseen Tasks from a Single Video Demonstrationcited by 61
- 2018Learning to Navigate the Webcited by 60
- 2018Hierarchical Reinforcement Learning for Zero-shot Generalization with Subtask Dependenciescited by 43
- 2016Learning to Push by Grasping: Using multiple tasks for effective learningcited by 37
- 2019DOM-Q-NET: Grounded RL on Structured Languagecited by 25
- 2020Meta Reinforcement Learning with Autonomous Inference of Subtask Dependenciescited by 25
- 2017A Deep Compositional Framework for Human-like Language Acquisition in Virtual Environmentcited by 3
- 2017Autonomous Extracting a Hierarchical Structure of Tasks in Reinforcement Learning and Multi-task Reinforcement Learningcited by 2
Built on by
Citing papers per year: 2023: 3.
3 papers.