English

Neural Networks Use Distance Metrics

Machine Learning 2024-11-28 v1 Artificial Intelligence Machine Learning

Abstract

We present empirical evidence that neural networks with ReLU and Absolute Value activations learn distance-based representations. We independently manipulate both distance and intensity properties of internal activations in trained models, finding that both architectures are highly sensitive to small distance-based perturbations while maintaining robust performance under large intensity-based perturbations. These findings challenge the prevailing intensity-based interpretation of neural network activations and offer new insights into their learning and decision-making processes.

Keywords

Cite

@article{arxiv.2411.17932,
  title  = {Neural Networks Use Distance Metrics},
  author = {Alan Oursland},
  journal= {arXiv preprint arXiv:2411.17932},
  year   = {2024}
}

Comments

8 pages excluding references and appendix. 12 pages total. 3 figures. The code for the experiments in this paper is available at https://github.com/alanoursland/neural_networks_use_distance_metrics

R2 v1 2026-06-28T20:13:54.199Z