Learning Parameterized Task Structure for Generalization to Unseen Entities
Not on the map. Why?
Abstract (arXiv)
Real world tasks are hierarchical and compositional. Tasks can be composed of multiple subtasks (or sub-goals) that are dependent on each other. These subtasks are defined in terms of entities (e.g., "apple", "pear") that can be recombined to form new subtasks (e.g., "pickup apple", and "pickup pear"). To solve these tasks efficiently, an agent must infer subtask dependencies (e.g. an agent must execute "pickup apple" before "place apple in pot"), and generalize the inferred dependencies to new subtasks (e.g. "place apple in pot" is similar to "place apple in pan"). Moreover, an agent may also need to solve unseen tasks, which can involve unseen entities. To this end, we formulate parameterized subtask graph inference (PSGI), a method for modeling subtask dependencies using first-order logic with subtask entities. To facilitate this, we learn entity attributes in a zero-shot manner, which are used as quantifiers (e.g. "is_pickable(X)") for the parameterized subtask graph. We show this approach accurately learns the latent structure on hierarchical and compositional tasks more efficiently than prior work, and show PSGI can generalize by modelling structure on subtasks unseen during adaptation.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 14 of its 27 references that are in the map. First posted on arXiv 28 Mar 2022.
Builds on
14 of 27 references.
References · 14
- 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
- 2015Effective Approaches to Attention-based Neural Machine Translationcited by 1,024
- 2017AI2-THOR: An Interactive 3D Environment for Visual AIcited by 850
- 2016RL2: Fast Reinforcement Learning via Slow Reinforcement Learningcited by 490
- 2016Modular Multitask Reinforcement Learning with Policy Sketchescited by 200
- 2017Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learningcited by 97
- 2017Neural Task Programming: Learning to Generalize Across Hierarchical Taskscited by 75
- 2018Neural Task Graphs: Generalizing to Unseen Tasks from a Single Video Demonstrationcited by 60
- 2018Rearranging the Familiar: Testing Compositional Generalization in Recurrent Networkscited by 51
- 2018Hierarchical Reinforcement Learning for Zero-shot Generalization with Subtask Dependenciescited by 43
- 2020Meta Reinforcement Learning with Autonomous Inference of Subtask Dependenciescited by 25
- 2020Reinforcement Learning for Sparse-Reward Object-Interaction Tasks in a First-person Simulated 3D Environmentcited by 8
Built on by
Citing papers per year: 2023: 3.
3 papers.