English

Learning Modular Structures That Generalize Out-of-Distribution

Machine Learning 2022-08-09 v1 Artificial Intelligence

Abstract

Out-of-distribution (O.O.D.) generalization remains to be a key challenge for real-world machine learning systems. We describe a method for O.O.D. generalization that, through training, encourages models to only preserve features in the network that are well reused across multiple training domains. Our method combines two complementary neuron-level regularizers with a probabilistic differentiable binary mask over the network, to extract a modular sub-network that achieves better O.O.D. performance than the original network. Preliminary evaluation on two benchmark datasets corroborates the promise of our method.

Keywords

Cite

@article{arxiv.2208.03753,
  title  = {Learning Modular Structures That Generalize Out-of-Distribution},
  author = {Arjun Ashok and Chaitanya Devaguptapu and Vineeth Balasubramanian},
  journal= {arXiv preprint arXiv:2208.03753},
  year   = {2022}
}

Comments

Accepted at AAAI 2022 Student Abstract and Poster Program

R2 v1 2026-06-25T01:32:57.407Z