Loading...

Reverse engineering recurrent networks for sentiment classification reveals line attractor dynamics

Recurrent neural networks (RNNs) are a widely used tool for modeling sequential data, yet they are often treated as inscrutable black boxes. Given a trained recurrent network, we would like to reverse engineer it–to obtain a quantitative, interpretable description of how it solves a particular task....

Fuld beskrivelse

Na minha lista:
Bibliografiske detaljer
Udgivet i:Adv Neural Inf Process Syst
Main Authors: Maheswaranathan, Niru, Williams, Alex H., Golub, Matthew D., Ganguli, Surya, Sussillo, David
Format: Artigo
Sprog:Inglês
Udgivet: 2019
Fag:
Online adgang:https://ncbi.nlm.nih.gov/pmc/articles/PMC7416638/
https://ncbi.nlm.nih.gov/pubmed/32782423
Tags: Tilføj Tag
Ingen Tags, Vær først til at tagge denne postø!