DS-2020-06: Hierarchy and interpretability in neural models of language processing

DS-2020-06: Hupkes, Dieuwke (2020) Hierarchy and interpretability in neural models of language processing. Doctoral thesis, University of Amsterdam.

[thumbnail of Full Text] Text (Full Text)
DS-2020-06.text.pdf

Download (11MB)
[thumbnail of Samenvatting] Text (Samenvatting)
DS-2020-06.samenvatting.txt

Download (2kB)
[thumbnail of Cover] Text (Cover)
DS-2020-06.cover.pdf

Download (17kB)

Abstract

Artificial neural networks have become remarkably successful on many natural language processing tasks. In this dissertation, I explore if these successes make them useful as explanatory models of human language processing, focusing in particular on hierarchical compositionality and recurrent neural networks (RNNs). I consider two questions:

- Are RNNs in fact capable of processing hierarchical compositional structures?
- How can we obtain insight in how they do so?

This dissertation is divided into three parts. In part one, I consider artificial languages, which provide a clean setup in which processing of structure can be studied in isolation. In this part, I also introduce diagnostic classification -- an interpretability technique that plays an important role in this dissertation -- and reflect upon what it means for a model to be able to process hierarchical compositionality.

In part two, I consider language models, trained on naturalistic data. Such models have been shown to capture syntax-sensitive long-distance subject-verb relationships. I investigate how they do so. I present detailed analyses of their inner dynamics, using diagnostic classification, neuron ablation and generalised contextual decomposition.

Lastly, in part three, I consider if a model's solution can be changed through an adapted learning signal.

In summary, in this dissertation I present many different analyses concerning the abilities of RNNs to process hierarchical structure, as well as several techniques to understand these blackbox models. The results sketch a positive picture of the usefulness of such models as explanatory models of processing languages with hierarchical compositional semantics.

Item Type: Thesis (Doctoral)
Report Nr: DS-2020-06
Series Name: ILLC Dissertation (DS) Series
Year: 2020
Subjects: Language
Depositing User: Dr Marco Vervoort
Date Deposited: 14 Jun 2022 15:17
Last Modified: 14 Jun 2022 15:17
URI: https://eprints.illc.uva.nl/id/eprint/2175

Actions (login required)

View Item View Item