Independent scholarship · Open accessThe monthly digest

Indexed text attachment

representation-theory-of-parametrized-neural-models-on-spaces.tex

Source for Representation Theory of Parametrized Neural Models on Spaces

View paper · Download original file

Lines 83–182 of 1343. Source text is displayed literally.

83\newcommand{\Z}{\mathbb Z}84\newcommand{\overB}{\E_{/B}}85\tikzcdset{every label/.append style={font=\small}}86\begin{document}87\twocolumn[88\begin{@twocolumnfalse}89\noindent\textcolor{paperaccent}{\rule{\textwidth}{1pt}}\par90\vspace{11pt}91{\noindent\fontsize{25}{30}\selectfont\bfseries92Representation Theory of Parametrized\\93Neural Models on Spaces\par}94\vspace{8pt}95{\noindent\large\color{papermuted}96Statistical families, equivariant layers, and homotopy-coherent models\par}97\vspace{10pt}98{\noindent\small November 2026\par}99\vspace{9pt}100\noindent\begin{minipage}{\textwidth}\small101\textbf{Abstract.}102A parametrized network over a space is a family of evaluation maps.103We represent its realizations and reparametrizations by a slice over an104internal mapping object, prove base change and descent, and extend the105construction to stochastic outputs for a strong probability monad.106On smooth charts this gives deterministic manifold networks and107conditional statistical models, together with pullback Fisher geometry.108For compact symmetries we classify bounded layers linear in a moving109parameter by multiplicity-space intertwiners. The classification exhibits110frequency shifts that cannot occur for a fixed equivariant layer.111For parameters with local symmetries we retain stabilizers and transition112data in a quotient stack. A neural decoder on a twisted real line illustrates why113global probability laws can lack a global smooth weight. Mapping spectra114supply a stable version of equivariance with higher intertwiners.115\par\medskip116\noindent\textbf{Keywords.}\enspace117Parametrized networks, representation theory, equivariance, Fisher information,118descent, quotient stacks.119\end{minipage}120\vspace{17pt}121\end{@twocolumnfalse}122]123\thispagestyle{plain}124 125\section{From statistical families to representations}126 127A supervised model commonly starts with a parameter manifold $P$, a128feature space $X$, and a map $f(p,x)$ valued in labels, predictions, or129probability distributions. A convolutional network has additional130structure. Features may live in bundles on an observation space, and a131symmetry can transport both features and weights. Two local weights can132also present the same model via a change of frame. These cases require133tracking evaluation, transport, and equivalences of presentations.134 135\emph{Representation Theory in Higher Categories} develops ordinary136and coherent group actions,137induced feature spaces, and equivariant convolutions138\cite[Chs.~17, 20, 27--28]{reprbook}. We start with its idea of encoding139actions by functors and relate evaluation, symmetry, and descent.140Evaluation records what a parameter computes. Representation theory141controls how evaluations transform. Descent determines when local142evaluations and parameter presentations assemble across a base.143Theorem~\ref{thm:relative-representation} gives the general comparison.144Theorem~\ref{thm:harmonic} computes the moving-parameter part in a145compact linear setting.146 147Parametrized composition and supervised learning have categorical148formulations through learners, lenses, and reverse differentiation149\cite{fong-spivak-tuyeras,cruttwell}. Markov categories organize channels150and sufficient statistics \cite{fritz-2020}, while equivariant151convolution on homogeneous spaces is described by induced152representations \cite{cohen-geiger-weiler}. Here we apply representation153constraints to the evaluation families themselves. This permits the same154typing of a manifold layer, a statistical decoder, and a locally presented155system of layers.156 157\section{Relative parametrized models}158\label{sec:relative}159 160Let $\E$ be an $\infty$-topos and $B$ an object of $\E$.161Possible choices are the $\infty$-category $\Spc$ of spaces and the162$\infty$-topos of sheaves of spaces on a smooth site. Work first163in $\overB$. For $X,Y$ over $B$, write $[X,Y]_B$ for their164exponential in that slice. Its evaluation map is165$[X,Y]_B\times_B X\to Y$ over $B$. A choice of site specifies which166families of sections are admitted. Smooth manifolds will be embedded167by their smooth functors of points.168 169\begin{definition}[A parametrized layer]170A layer from $X$ to $Y$ over $B$ consists of $P\to B$ and an171evaluation $f:P\times_B X\to Y$. A map to $(Q,g)$ is a morphism172$u:P\to Q$ over $B$ with a specified homotopy173$f\simeq g(u\times_B\id_X)$ over $B$. Higher morphisms are the174compatibilities in the mapping space of the resulting slice.175\end{definition}176 177Write $r_f:P\to[X,Y]_B$ for the transpose of $f$. Given a second178layer $(Q,g):Y\to Z$, the series composite has parameter179$P\times_B Q$ and evaluation $(p,q,x)\mapsto g(q,f(p,x))$.180Here and below parametric formulas describe generalized points in the181slice. A shared parameter $P$ can be imposed by reparametrization along182the diagonal $P\to P\times_B P$. More general tying is expressed by a