Manifolds and Modules: How Function Develops in a Neural Foundation Model

TL;DR

This paper investigates a state-of-the-art neural foundation model's internal representations, revealing how different processing stages encode stimuli and how recurrent modules enhance temporal discrimination, providing insights into biological neural processes.

Contribution

It introduces a novel analysis of neural foundation models by characterizing neurons' response properties and manifold structures, bridging AI models and biological neural systems.

Findings

Recurrent modules improve temporal stimulus separation.

Different model stages exhibit distinct representational structures.

Specialized feature maps in the readout enhance biological fidelity.

Abstract

Foundation models have shown remarkable success in fitting biological visual systems; however, their black-box nature inherently limits their utility for understanding brain function. Here, we peek inside a SOTA foundation model of neural activity (Wang et al., 2025) as a physiologist might, characterizing each 'neuron' based on its temporal response properties to parametric stimuli. We analyze how different stimuli are represented in neural activity space by building decoding manifolds, and we analyze how different neurons are represented in stimulus-response space by building neural encoding manifolds. We find that the different processing stages of the model (i.e., the feedforward encoder, recurrent, and readout modules) each exhibit qualitatively different representational structures in these manifolds. The recurrent module shows a jump in capabilities over the encoder module by 'pushing apart' the representations of different temporal stimulus patterns; while the readout module achieves biological fidelity by using numerous specialized feature maps rather than biologically plausible mechanisms. Overall, we present this work as a study of the inner workings of a prominent neural foundation model, gaining insights into the biological relevance of its internals through the novel analysis of its neurons' joint temporal response patterns.