English

Making Sense Of Distributed Representations With Activation Spectroscopy

Machine Learning 2025-01-28 v1 Computer Vision and Pattern Recognition

Abstract

In the study of neural network interpretability, there is growing evidence to suggest that relevant features are encoded across many neurons in a distributed fashion. Making sense of these distributed representations without knowledge of the network's encoding strategy is a combinatorial task that is not guaranteed to be tractable. This work explores one feasible path to both detecting and tracing the joint influence of neurons in a distributed representation. We term this approach Activation Spectroscopy (ActSpec), owing to its analysis of the pseudo-Boolean Fourier spectrum defined over the activation patterns of a network layer. The sub-network defined between a given layer and an output logit is cast as a special class of pseudo-Boolean function. The contributions of each subset of neurons in the specified layer can be quantified through the function's Fourier coefficients. We propose a combinatorial optimization procedure to search for Fourier coefficients that are simultaneously high-valued, and non-redundant. This procedure can be viewed as an extension of the Goldreich-Levin algorithm which incorporates additional problem-specific constraints. The resulting coefficients specify a collection of subsets, which are used to test the degree to which a representation is distributed. We verify our approach in a number of synthetic settings and compare against existing interpretability benchmarks. We conclude with a number of experimental evaluations on an MNIST classifier, and a transformer-based network for sentiment analysis.

Keywords

Cite

@article{arxiv.2501.15435,
  title  = {Making Sense Of Distributed Representations With Activation Spectroscopy},
  author = {Kyle Reing and Greg Ver Steeg and Aram Galstyan},
  journal= {arXiv preprint arXiv:2501.15435},
  year   = {2025}
}
R2 v1 2026-06-28T21:18:03.494Z