Exact capacity of the \emph{wide} hidden layer treelike neural networks with generic activations

TL;DR

This paper derives explicit capacity formulas for wide treelike neural networks with generic activations using advanced Random Duality Theory, confirming results with statistical physics methods and demonstrating rapid convergence of the lifting approach.

Contribution

It introduces a fully lifted RDT framework to precisely analyze the capacity of wide treelike neural networks with various activations, simplifying numerical evaluations.

Findings

Explicit capacity formulas for ReLU, quadratic, erf, and tanh activations.

Rapid convergence of the lifting method within three levels.

Capacity results match those from statistical physics approaches.

Abstract

Recent progress in studying \emph{treelike committee machines} (TCM) neural networks (NN) in \cite{Stojnictcmspnncaprdt23,Stojnictcmspnncapliftedrdt23,Stojnictcmspnncapdiffactrdt23} showed that the Random Duality Theory (RDT) and its a \emph{partially lifted}(pl RDT) variant are powerful tools that can be used for very precise networks capacity analysis. Here, we consider \emph{wide} hidden layer networks and uncover that certain aspects of numerical difficulties faced in \cite{Stojnictcmspnncapdiffactrdt23} miraculously disappear. In particular, we employ recently developed \emph{fully lifted} (fl) RDT to characterize the \emph{wide} ($d\rightarrow \infty$) TCM nets capacity. We obtain explicit, closed form, capacity characterizations for a very generic class of the hidden layer activations. While the utilized approach significantly lowers the amount of the needed numerical evaluations, the ultimate fl RDT usefulness and success still require a solid portion of the residual numerical work. To get the concrete capacity values, we take four very famous activations examples: \emph{\textbf{ReLU}}, \textbf{\emph{quadratic}}, \textbf{\emph{erf}}, and \textbf{\emph{tanh}}. After successfully conducting all the residual numerical work for all of them, we uncover that the whole lifting mechanism exhibits a remarkably rapid convergence with the relative improvements no better than $\sim 0.1\%$ happening already on the 3-rd level of lifting. As a convenient bonus, we also uncover that the capacity characterizations obtained on the first and second level of lifting precisely match those obtained through the statistical physics replica theory methods in \cite{ZavPeh21} for the generic and in \cite{BalMalZech19} for the ReLU activations.