Discovering modular solutions that generalize compositionally

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Schug, Simon, Kobayashi, Seijin, Akram, Yassir, Wołczyk, Maciej, Proca, Alexandra, von Oswald, Johannes, Pascanu, Razvan, Sacramento, João, Steger, Angelika
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917621360754688
author Schug, Simon
Kobayashi, Seijin
Akram, Yassir
Wołczyk, Maciej
Proca, Alexandra
von Oswald, Johannes
Pascanu, Razvan
Sacramento, João
Steger, Angelika
author_facet Schug, Simon
Kobayashi, Seijin
Akram, Yassir
Wołczyk, Maciej
Proca, Alexandra
von Oswald, Johannes
Pascanu, Razvan
Sacramento, João
Steger, Angelika
contents Many complex tasks can be decomposed into simpler, independent parts. Discovering such underlying compositional structure has the potential to enable compositional generalization. Despite progress, our most powerful systems struggle to compose flexibly. It therefore seems natural to make models more modular to help capture the compositional nature of many tasks. However, it is unclear under which circumstances modular systems can discover hidden compositional structure. To shed light on this question, we study a teacher-student setting with a modular teacher where we have full control over the composition of ground truth modules. This allows us to relate the problem of compositional generalization to that of identification of the underlying modules. In particular we study modularity in hypernetworks representing a general class of multiplicative interactions. We show theoretically that identification up to linear transformation purely from demonstrations is possible without having to learn an exponential number of module combinations. We further demonstrate empirically that under the theoretically identified conditions, meta-learning from finite data can discover modular policies that generalize compositionally in a number of complex environments.
format Preprint
id arxiv_https___arxiv_org_abs_2312_15001
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Discovering modular solutions that generalize compositionally
Schug, Simon
Kobayashi, Seijin
Akram, Yassir
Wołczyk, Maciej
Proca, Alexandra
von Oswald, Johannes
Pascanu, Razvan
Sacramento, João
Steger, Angelika
Machine Learning
Neural and Evolutionary Computing
Many complex tasks can be decomposed into simpler, independent parts. Discovering such underlying compositional structure has the potential to enable compositional generalization. Despite progress, our most powerful systems struggle to compose flexibly. It therefore seems natural to make models more modular to help capture the compositional nature of many tasks. However, it is unclear under which circumstances modular systems can discover hidden compositional structure. To shed light on this question, we study a teacher-student setting with a modular teacher where we have full control over the composition of ground truth modules. This allows us to relate the problem of compositional generalization to that of identification of the underlying modules. In particular we study modularity in hypernetworks representing a general class of multiplicative interactions. We show theoretically that identification up to linear transformation purely from demonstrations is possible without having to learn an exponential number of module combinations. We further demonstrate empirically that under the theoretically identified conditions, meta-learning from finite data can discover modular policies that generalize compositionally in a number of complex environments.
title Discovering modular solutions that generalize compositionally
topic Machine Learning
Neural and Evolutionary Computing
url https://arxiv.org/abs/2312.15001