AI Chat AI Image Generator AI Video Text to Speech

The inverse problem for neural networks

08/27/2023

∙

by Marcelo Forets, et al.

∙

∙

We study the problem of computing the preimage of a set under a neural network with piecewise-affine activation functions. We recall an old result that the preimage of a polyhedral set is again a union of polyhedral sets and can be effectively computed. We show several applications of computing the preimage for analysis and interpretability of neural networks.

Marcelo Forets
6 publications
Christian Schilling
13 publications

research

∙ 01/30/2023

Formalizing Piecewise Affine Activation Functions of Neural Networks in Coq

Verification of neural networks relies on activation functions being pie...

0 Andrei Aleksandrov, et al. ∙

research

∙ 10/09/2019

Dissecting Deep Neural Networks

In exchange for large quantities of data and processing power, deep neur...

0 Haakon Robinson, et al. ∙

research

∙ 05/19/2022

Minimal Explanations for Neural Network Predictions

Explaining neural network predictions is known to be a challenging probl...

0 Ouns El Harzli, et al. ∙

research

∙ 01/22/2019

On Connected Sublevel Sets in Deep Learning

We study sublevel sets of the loss function in training deep neural netw...

0 Quynh Nguyen, et al. ∙

research

∙ 02/13/2022

Learning from Randomly Initialized Neural Network Features

We present the surprising result that randomly initialized neural networ...

0 Ehsan Amid, et al. ∙

research

∙ 09/04/2017

Optimal deep neural networks for sparse recovery via Laplace techniques

This paper introduces Laplace techniques for designing a neural network,...

0 Steffen Limmer, et al. ∙

research

∙ 03/05/2022

Tabula: Efficiently Computing Nonlinear Activation Functions for Secure Neural Network Inference

Multiparty computation approaches to secure neural network inference tra...

0 Maximilian Lam, et al. ∙