# Refactoring Policy for Compositional Generalizability using Self-Supervised Object Proposals

arXiv: https://arxiv.org/abs/2011.00971 · HTML: https://mlsanity.com/paper/2011.00971/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 26 Oct 2020.

## Built on by (most cited first, 9 of 9)

- [UniDexGrasp++: Improving Dexterous Grasping Policy Learning via Geometry-aware Curriculum and Iterative Generalist-Specialist Learning](https://mlsanity.com/paper/2304.00464/) (2023)
- [A Consciousness-Inspired Planning Agent for Model-Based Reinforcement Learning](https://mlsanity.com/paper/2106.02097/) (2021)
- [How to Reuse and Compose Knowledge for a Lifetime of Tasks: A Survey on Continual Learning and Functional Composition](https://mlsanity.com/paper/2207.07730/) (2022)
- [Improving Policy Optimization with Generalist-Specialist Learning](https://mlsanity.com/paper/2206.12984/) (2022)
- [Chain-of-Thought Predictive Control](https://mlsanity.com/paper/2304.00776/) (2023)
- [PhysReaction: Physically Plausible Real-Time Humanoid Reaction Synthesis via Forward Dynamics Guided 4D Imitation](https://mlsanity.com/paper/2404.01081/) (2024)
- [Learning to Design and Construct Bridge without Blueprint](https://mlsanity.com/paper/2108.02439/) (2021)
- [Robotic Manipulation Datasets for Offline Compositional Reinforcement Learning](https://mlsanity.com/paper/2307.07091/) (2023)
- [Learning Two-Step Hybrid Policy for Graph-Based Interpretable Reinforcement Learning](https://mlsanity.com/paper/2201.08520/) (2022)

## Abstract (arXiv)

We study how to learn a policy with compositional generalizability. We propose a two-stage framework, which refactorizes a high-reward teacher policy into a generalizable student policy with strong inductive bias. Particularly, we implement an object-centric GNN-based student policy, whose input objects are learned from images through self-supervised learning. Empirically, we evaluate our approach on four difficult tasks that require compositional generalizability, and achieve superior performance compared to baselines.