English

One Network Fits All? Modular versus Monolithic Task Formulations in Neural Networks

Machine Learning 2021-03-30 v1 Artificial Intelligence Machine Learning

Abstract

Can deep learning solve multiple tasks simultaneously, even when they are unrelated and very different? We investigate how the representations of the underlying tasks affect the ability of a single neural network to learn them jointly. We present theoretical and empirical findings that a single neural network is capable of simultaneously learning multiple tasks from a combined data set, for a variety of methods for representing tasks -- for example, when the distinct tasks are encoded by well-separated clusters or decision trees over certain task-code attributes. More concretely, we present a novel analysis that shows that families of simple programming-like constructs for the codes encoding the tasks are learnable by two-layer neural networks with standard training. We study more generally how the complexity of learning such combined tasks grows with the complexity of the task codes; we find that combining many tasks may incur a sample complexity penalty, even though the individual tasks are easy to learn. We provide empirical support for the usefulness of the learning bounds by training networks on clusters, decision trees, and SQL-style aggregation.

Keywords

Cite

@article{arxiv.2103.15261,
  title  = {One Network Fits All? Modular versus Monolithic Task Formulations in Neural Networks},
  author = {Atish Agarwala and Abhimanyu Das and Brendan Juba and Rina Panigrahy and Vatsal Sharan and Xin Wang and Qiuyi Zhang},
  journal= {arXiv preprint arXiv:2103.15261},
  year   = {2021}
}

Comments

30 pages, 6 figures

R2 v1 2026-06-24T00:37:52.454Z