Compositional Interfaces for Compositional Generalization
Abstract
In this work, we study the effectiveness of a modular architecture for compositional generalization and transfer learning in the embodied agent setting. We develop an environment that allows us to independently vary perceptual modalities and action and task instructions, and use it to carefully analyze the agent's performance in these compositions. Our experiments demonstrate strong zero-shot performance on held-out combinations of perception, action, and instruction spaces; as well as fast adaptation to new perceptual spaces without the loss of performance.
Cite
Text
Luketina et al. "Compositional Interfaces for Compositional Generalization." ICML 2023 Workshops: ES-FoMO, 2023.Markdown
[Luketina et al. "Compositional Interfaces for Compositional Generalization." ICML 2023 Workshops: ES-FoMO, 2023.](https://mlanthology.org/icmlw/2023/luketina2023icmlw-compositional/)BibTeX
@inproceedings{luketina2023icmlw-compositional,
title = {{Compositional Interfaces for Compositional Generalization}},
author = {Luketina, Jelena and Lanchantin, Jack and Sukhbaatar, Sainbayar and Szlam, Arthur},
booktitle = {ICML 2023 Workshops: ES-FoMO},
year = {2023},
url = {https://mlanthology.org/icmlw/2023/luketina2023icmlw-compositional/}
}