Searcharxiv⌕ Search

arXiv · cond-mat/0107587

Replica symmetric evaluation of the information transfer in a two-layer network in presence of continuous+discrete stimuli

Abstract

In a previous report we have evaluated analytically the mutual information between the firing rates of N independent units and a set of multi-dimensional continuous+discrete stimuli, for a finite population size and in the limit of large noise. Here, we extend the analysis to the case of two interconnected populations, where input units activate output ones via gaussian weights and a threshold linear transfer function. We evaluate the information carried by a population of M output units, again about continuous+discrete correlates. The mutual information is evaluated solving saddle point equations under the assumption of replica symmetry, a method which, by taking into account only the term linear in N of the input information, is equivalent to assuming the noise to be large. Within this limitation, we analyze the dependence of the information on the ratio M/N, on the selectivity of the input units and on the level of the output noise. We show analytically, and confirm numerically, that in the limit of a linear transfer function and of a small ratio between output and input noise, the output information approaches asymptotically the information carried in input. Finally, we show that the information loss in output does not depend much on the structure of the stimulus, whether purely continuous, purely discrete or mixed, but only on the position of the threshold nonlinearity, and on the ratio between input and output noise.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Valeria Del Prete, Alessandro Treves. 2001-07-30. Replica symmetric evaluation of the information transfer in a two-layer network in presence of continuous+discrete stimuli. https://arxiv.org/abs/cond-mat/0107587

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

Embedding stochastic dynamics in latent manifolds of recurrent neural networks

Recurrent neural networks (RNNs) provide a theoretical framework for understanding computation in biological neural circuits through their ability to support rich time-dependent behaviour. Whilst high-dimensional by construction, they can be used to encode low-dimensional dynamical systems, for example by restricting to low-rank connectivity, hence mirroring the emergence of latent neural manifolds found in empirical neural data. Here, we exploit a relationship between RNNs and a class of universal approximators to train networks that can represent arbitrary, nonlinear stochastic differential equations (SDEs), with given drift and diffusion coefficients, within a low-dimensional latent subspace. We show that RNNs can faithfully embed the drift and diffusion of a given SDE, including nonlinear and nonequilibrium dynamics such as chaotic attractors. We apply this approach by constructing RNN realisations of stochastic systems that transiently explore various attractors through both input-driven switching and autonomous transitions driven by nonequilibrium currents, which we interpret as models of associative and sequential (episodic) memory. To elucidate how these dynamics are encoded in the network, we introduce decompositions of the RNN based on its asymmetric connectivity and its time-irreversibility. Our results show that neural populations can implement a broad class of stochastic dynamics within low-dimensional manifolds, unifying ideas from the dynamics of neural computation, the theory of neural manifolds, associative memory, and nonequilibrium statistical mechanics.

cond-mat.dis-nn↗

Deep-layered machines have a built-in Occam's razor

Many complex systems map vast numbers of microscopic descriptions onto far fewer macroscopic outcomes. Surprisingly, these input-output maps strongly favor simple outputs. In an exactly solvable deep-layered machine, we prove that simplicity bias strengthens with every layer. The probability of an output function $F$ approaches $\Pr(F)\propto 2^{-K(F)}$, where $K$ is the Kolmogorov complexity. Our results show that hierarchical composition alone can generate an intrinsic Occam's razor, suggesting a general mechanism for simplicity bias in learning and evolution.

cond-mat.dis-nn↗

Cup and Cap Topological Neural Network

Topological Deep Learning (TDL) is designed to learn features associated not only with nodes but also with edges, triangles and higher dimensional simplices. However, most TDL approaches, being based on boundary operators and Hodge Laplacians, have the limitation that features and signals cannot be lifted or lowered across more than one dimension per layer. To overcome this limitation, in this work, we propose the adoption of the cup and cap products. Specifically, we formulate the Cup and Cap Topological Neural Network (CCNN), a topological deep learning architecture designed to learn node-based variables (or 0-cochains) by taking into account their many-body interactions (e.g. triangles) present in the data. We validate CCNN by assessing its performance on the TopoBench datasets, demonstrating its competitive performance with respect to other simplicial complex neural networks.

cond-mat.dis-nn↗