Only Brains Align with Brains: Cross-Region Alignment Patterns Expose Limits of Normative Models

Larissa Höfling, Matthias Tangemann, Lotta Piefke, Susanne Keller, Matthias Bethge, Katrin Franke

Keywords: brain alignment, benchmarking, representational similarity analysis, video models

TL;DR: We expose the limits of brain alignment of SOTA video models, and propose a framework based on cross-region alignment patterns in the brain towards more robust and meaningful assessment of brain-model alignment.

Abstract:

Neuroscientists and computer vision researchers use model–brain alignment benchmarks to compare artificial and biological vision systems. These benchmarks rank models according to alignment measures such as the similarity of representational geometry or the predictability of neural responses from model activations. However, recent works have identified a number of problems with these rankings, among them their lack of discriminative power and robustness, raising the conceptual question of what it means for a model to be 'brain-aligned'.
Here we introduce *alignment patterns* - characteristic functional relationship profiles of each brain region to all others - and propose that models should reproduce these patterns to qualify as brain-aligned.
First, we apply a standard benchmarking pipeline to a broad spectrum of vision models on the BOLD Moments video fMRI dataset across visual regions of interest (ROIs).
We find diverse models appear *equivalent* in their brain alignment, reflecting the lack of discriminative power of conventional alignment benchmarking pipelines based on pointwise, i.e. ROI-layer, comparisons.
In contrast, *alignment pattern analysis (APA)* is a second-order structural consistency test: a model aligned to a given ROI should reproduce that ROI’s characteristic cross-region alignment profile.
Applying APA, we find that, while these patterns are highly stable across brains of different subjects, even top-ranked models often fail to capture them. Notably, models that appear effectively equivalent in their pointwise alignment to an ROI diverge sharply under the relational criterion, demonstrating the added discriminative value of APA.
Finally, we argue for a clearer distinction between the criteria a model must meet to serve as a tool versus as a computational model for human visual cortex. Conventional alignment measures may be sufficient for identifying neurally predictive models, but claims about computational or algorithmic similarity may require a stronger basis of evidence, including the reproducibility of relational alignment patterns.

Primary Area: applications to neuroscience & cognitive science

Submission Number: 13969