analiza aiqbench
This research applies statistical physics frameworks to analyze two-layer Bayesian neural networks with extensive width scaling proportional to input dimensions. The study reveals critical learning transitions around the interpolation regime, demonstrating how feature specialization emerges as data availability increases and addressing gaps between theoretical Bayes-optimal limits and practical training dynamics.
Statistical mechanics of extensive-width Bayesian neural networks near interpolation
arXiv:2505.24849v2 Announce Type: replace-cross Abstract: For three decades statistical mechanics has been providing a framework to analyse neural networks. However, the theoretically tractable models, e.g., perceptrons, random features models and kernel machines, or multi-index models and committee machines with few neurons, remained simple compared to those used in applications. In this paper we help reducing the gap between practical networks and their theoretical understanding through a statistical physics analysis of the supervised learning of a two-layer fully connected network with generic weight distribution and activation function, whose hidden layer is large but remains proportional to the inputs dimension. This makes it more realistic than infinitely wide networks where no feature learning occurs, but also more expressive than narrow ones or with fixed inner weights. We focus on the Bayes-optimal learning in the teacher-student scenario, i.e., with a dataset generated by another network with the same architecture. We operate around interpolation, where the number of trainable parameters and of data are comparable and feature learning emerges. Our analysis uncovers a rich phenomenology with various learning transitions as the number of data increases. In particular, the more strongly the features (i.e., hidden neurons of the target) contribute to the observed responses, the less data is needed to learn them. Moreover, when the data is scarce, the model only learns non-linear combinations of the teacher weights, rather than "specialising" by aligning its weights with the teacher's. Specialisation occurs only when enough data becomes available, but it can be hard to find for practical training algorithms, possibly due to statistical-to-computational~gaps.
Abstract:For three decades statistical mechanics has been providing a framework to analyse neural networks. However, the theoretically tractable models, e.g., perceptrons, random features models and kernel machines, or multi-index models and committee machines with few neurons, remained simple compared to those used in applications. In this paper we help reducing the gap between practical networks and their theoretical understanding through a statistical physics analysis of the supervised learning of a two-layer fully connected network with generic weight distribution and activation function, whose hidden layer is large but remains proportional to the inputs dimension. This makes it more realistic than infinitely wide networks where no feature learning occurs, but also more expressive than narrow ones or with fixed inner weights. We focus on the Bayes-optimal learning in the teacher-student scenario, i.e., with a dataset generated by another network with the same architecture. We operate around interpolation, where the number of trainable parameters and of data are comparable and feature learning emerges. Our analysis uncovers a rich phenomenology with various learning transitions as the number of data increases. In particular, the more strongly the features (i.e., hidden neurons of the target) contribute to the observed responses, the less data is needed to learn them. Moreover, when the data is scarce, the model only learns non-linear combinations of the teacher weights, rather than "specialising" by aligning its weights with the teacher's. Specialisation occurs only when enough data becomes available, but it can be hard to find for practical training algorithms, possibly due to statistical-to-computational~gaps.
This submission has been superseded by arXiv:2510.24616, which substantially improves the presentation of the theory devised here for the shallow case, adds more numerical evidence, pushes forward the approach to deeper targets/models, and has been published in Phys. Rev. X 16, 031014 (2026)
Machine Learning (stat.ML); Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistical Mechanics (cond-mat.stat-mech); Information Theory (cs.IT); Machine Learning (cs.LG)
arXiv:2505.24849 [stat.ML]
(or arXiv:2505.24849v2 [stat.ML] for this version)
https://doi.org/10.1103/56sb-pdh6 DOI(s) linking to related resources
Text integral sub licența deschisă a publicației · 326 cuvinte
