Improving Systematic Generalization Through Modularity and Augmentation

Laura Ruis, Brenden M. Lake

Research output: Contribution to conferencePaperpeer-review


Systematic generalization is the ability to combine known parts into novel meaning; an important aspect of efficient human learning, but a weakness of neural network learning. In this work, we investigate how two well-known modeling principles-modularity and data augmentation-affect systematic generalization of neural networks in grounded language learning. We analyze how large the vocabulary needs to be to achieve systematic generalization and how similar the augmented data needs to be to the problem at hand. Our findings show that even in the controlled setting of a synthetic benchmark, achieving systematic generalization remains very difficult. After training on an augmented dataset with almost forty times more adverbs than the original problem, a non-modular baseline is not able to systematically generalize to a novel combination of a known verb and adverb. When separating the task into cognitive processes like perception and navigation, a modular neural network is able to utilize the augmented data and generalize more systematically, achieving 70% and 40% exact match increase over state-of-the-art on two gSCAN tests that have not previously been improved. We hope that this work gives insight into the drivers of systematic generalization, and what we still need to improve for neural networks to learn more like humans do.

Original languageEnglish (US)
Number of pages9
StatePublished - 2022
Event44th Annual Meeting of the Cognitive Science Society: Cognitive Diversity, CogSci 2022 - Toronto, Canada
Duration: Jul 27 2022Jul 30 2022


Conference44th Annual Meeting of the Cognitive Science Society: Cognitive Diversity, CogSci 2022


  • Data Augmentation
  • Modularity
  • Systematic Generalization

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Human-Computer Interaction
  • Cognitive Neuroscience


Dive into the research topics of 'Improving Systematic Generalization Through Modularity and Augmentation'. Together they form a unique fingerprint.

Cite this