CORTEXA
← Browse
arxivcs.LG2026-07-21

Do Sheaf Neural Networks Use Holonomy? A Measure--Intervene--Control Study

Ankit Grover, Rémi Bourgerie

Geometric architectures are often justified by internal mechanisms such as rotations, yet task performance alone cannot show whether those mechanisms drive predictions. Using sheaf neural networks (SNNs) as a testbed, we introduce the first basis-independent measurement of trained triangle-loop products, separating rotation, stalk-space area, and orientation. In a custom high-homophily GraphUniverse regime, Neural Sheaf Propagation (NSP) increases the triangle-weighted mean two-dimensional SO(2) loop rotation from 0.010 to 0.388 radians for triangle counting, while the community-detection comparison ends at 0.029 radians. Across the training-set-size experiment, replacing all learned SO(2) transports with identities sharply increases test error, establishing post-training sensitivity to the complete learned connection. However, a graph-summary ridge predictor is more accurate, diagonal maps also improve, and fixed-degree graphs develop increasing rotation without outperforming the training-mean predictor. This measure-intervene-control study separates geometric change, connection sensitivity, and evidence for triangle-specific computation.

View free PDFSource page

Related papers

arxivmath.LOcs.LGmath.FA2026-06-29

Fast approximation and learning of binary classification tasks in o-minimal structures using ReLU neural networks

Clemens Kinn, Philipp Petersen

We study binary classification problems whose decision sets are given by definable sets in o-minimal expansions of the real field. Motivated by cell decomposition of definable sets, we introduce traceable sets as a classical proxy for definable decision regions and analyze their…

View free PDFSource page
arxivcs.LG2026-06-29

Improved Predictive Performance and Interpretability for Mesomorphic Neural Networks Using Local Fidelity Regularization

Hugo L. Hammer, Vajira Thambawita, Kristoffer Herland Hellton, Pål Halvorsen

Interpretable Mesomorphic Neural Networks (IMNs) offer a promising framework that combines the predictive power of deep neural networks with the interpretability of linear models. However, the original formulation lacks safeguards to ensure that the learned interpretations are in…

View free PDFSource page
arxivcs.LGcs.AImath.DS2026-07-05

Empirical Minimal-Realisation Compression of Deep Neural Networks via Controllability-Observability Tests

Anis Hamadouche, Amir Hussain

Deep neural networks often contain substantial hidden-state redundancy, but most compression methods operate directly on weights, neurons, or quantised representations without explicitly characterising the dynamical role of internal states. This paper proposes a controllability-o…

View free PDFSource page