Subspace Chronicles: How Linguistic Information Emerges, Shifts and Interacts during Language Model Training

Research output: Conference Article in Proceeding or Book/Report chapterArticle in proceedingsResearchpeer-review

Abstract

Representational spaces learned via language modeling are fundamental to Natural Language Processing (NLP), however there has been limited understanding regarding how and when during training various types of linguistic information emerge and interact. Leveraging a novel information theoretic probing suite, which enables direct comparisons of not just task performance, but their representational subspaces, we analyze nine tasks covering syntax, semantics and reasoning, across 2M pre-training steps and five seeds. We identify critical learning phases across tasks and time, during which subspaces emerge, share information, and later disentangle to specialize. Across these phases, syntactic knowledge is acquired rapidly after 0.5% of full training. Continued performance improvements primarily stem from the acquisition of open-domain knowledge, while semantics and reasoning tasks benefit from later boosts to long-range contextualization and higher specialization. Measuring cross-task similarity further reveals that linguistically related tasks share information throughout training, and do so more during the critical phase of learning than before or after. Our findings have implications for model interpretability, multi-task learning, and learning from limited data.
Original languageEnglish
Title of host publicationFindings of the Association for Computational Linguistics: EMNLP 2023
Place of PublicationSingapore
PublisherAssociation for Computational Linguistics
Publication date6 Dec 2023
Pages13190-13208
Publication statusPublished - 6 Dec 2023

Fingerprint

Dive into the research topics of 'Subspace Chronicles: How Linguistic Information Emerges, Shifts and Interacts during Language Model Training'. Together they form a unique fingerprint.

Cite this