Basic syntax from speech: Spontaneous concatenation in unsupervised deep neural networks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Beguš, Gašper, Lu, Thomas, Wang, Zili
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913055762284544
author Beguš, Gašper
Lu, Thomas
Wang, Zili
author_facet Beguš, Gašper
Lu, Thomas
Wang, Zili
contents Computational models of syntax are predominantly text-based. Here we propose that the most basic first step in the evolution of syntax can be modeled directly from raw speech in a fully unsupervised way. We focus on one of the most ubiquitous and elementary suboperations of syntax -- concatenation. We introduce \textit{spontaneous concatenation}: a phenomenon where a ciwGAN/fiwGAN models (based on convolutional neural networks) trained on acoustic recordings of individual words start generating outputs with two or even three words concatenated without ever accessing data with multiple words in the training data. We replicate this finding in several independently trained models with different hyperparameters and training data. Additionally, networks trained on two words learn to embed words into novel unobserved word combinations. We also show that the concatenated outputs contain precursors to compositionality. To our knowledge, this is a previously unreported property of CNNs trained in the ciwGAN/fiwGAN setting on raw speech and has implications both for our understanding of how these architectures learn as well as for modeling syntax and its evolution in the brain from raw acoustic inputs. We also propose and formalize a neural mechanism called \textit{disinhibition} that outlines a possible artificial and biological neural pathway towards concatenation and compositionality and suggests our modeling is useful for generating testable predictions for biological and artificial neural processing of spoken language.
format Preprint
id arxiv_https___arxiv_org_abs_2305_01626
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Basic syntax from speech: Spontaneous concatenation in unsupervised deep neural networks
Beguš, Gašper
Lu, Thomas
Wang, Zili
Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
Computational models of syntax are predominantly text-based. Here we propose that the most basic first step in the evolution of syntax can be modeled directly from raw speech in a fully unsupervised way. We focus on one of the most ubiquitous and elementary suboperations of syntax -- concatenation. We introduce \textit{spontaneous concatenation}: a phenomenon where a ciwGAN/fiwGAN models (based on convolutional neural networks) trained on acoustic recordings of individual words start generating outputs with two or even three words concatenated without ever accessing data with multiple words in the training data. We replicate this finding in several independently trained models with different hyperparameters and training data. Additionally, networks trained on two words learn to embed words into novel unobserved word combinations. We also show that the concatenated outputs contain precursors to compositionality. To our knowledge, this is a previously unreported property of CNNs trained in the ciwGAN/fiwGAN setting on raw speech and has implications both for our understanding of how these architectures learn as well as for modeling syntax and its evolution in the brain from raw acoustic inputs. We also propose and formalize a neural mechanism called \textit{disinhibition} that outlines a possible artificial and biological neural pathway towards concatenation and compositionality and suggests our modeling is useful for generating testable predictions for biological and artificial neural processing of spoken language.
title Basic syntax from speech: Spontaneous concatenation in unsupervised deep neural networks
topic Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2305.01626