Expand Neurons, Not Parameters

Linghao Kong, Inimai Subramanian, Yonadav G Shavit, Micah Adler, Dan Alistarh, Nir N Shavit

Keywords: Interpretability, Feature Coding, Superposition, Sparsity

TL;DR: We show that taking a model and increasing its number of neurons without changing the number of non-zero parameters improves accuracy via a reduction in superposition

Abstract: This work demonstrates how increasing the number of neurons in a network without increasing its number of non-zero parameters improves performance. We show that this gain corresponds with a decrease in interference between multiple features that would otherwise share the same neurons. On symbolic tasks, specifically Boolean code problems, splitting each neuron into sparser sub-neurons with knowledge of the clauses systematically reduces polysemanticity metrics and yields higher task accuracy. Notably, even random splits of neuron weights approximate these gains, indicating that reduced collisions, not precise assignment, are a primary driver. Consistent with the superposition hypothesis, the benefits of this framework grow with increasing interference: when polysemantic load is high, accuracy improvements are the largest. Transferring these insights to real models—classifiers over CLIP embeddings, CNNs, and deeper multilayer networks—we find that widening networks while maintaining a constant non-zero parameter count consistently increases accuracy. These results identify an interpretability-grounded mechanism to leverage width against superposition, improving performance without increasing the number of non-zero parameters. Such a direction is well matched to modern accelerators, where memory movement of non-zero parameters, rather than raw compute, is often the dominant bottleneck.

Primary Area: interpretability and explainable AI

Submission Number: 23416