At this week’s MCQLL meeting, Gaurav Kamath will be presenting What do Language Models Capture About Scope Ambiguities? An abstract follows.
We will be meeting this Tuesday, March 7 at 3:00 PM. Meetings will be held both in person in room 117 of the McGill Linguistics department at 1085 Dr Penfield and on Zoom here.
Title: What do Language Models Capture About Scope Ambiguities?
Abstract: In recent years, language models (LMs) have achieved human-like performance on a range of tasks designed to test natural language understanding (NLU). Their high performance on NLU tasks has led to claims of these models ‘understanding’ natural language; as a result of the methods by which they are trained, however, little is known about the actual linguistic structures they induce. In this talk, I present ongoing work aimed at investigating the manner in which LMs capture a specific semantic phenomenon: scope ambiguity. Focusing on scope ambiguities generated by quantifiers, negation, modality, quantificational adverbs and intensional verbs, I will cover the results of an experiment conducted on language models and humans, and discuss future experiments aimed at answering the same question.