Implicit causality in GPT-2: a case study
It addresses the problem of evaluating language model reasoning for linguists and AI researchers, but is incremental as it builds on prior work.
This case study investigated whether GPT-2 captures human intuitions about implicit causality in sentence completion, finding it shows lower surprisal for congruent pronouns and that verb frequency affects performance, while also developing a method to avoid bias in human ratings.
This case study investigates the extent to which a language model (GPT-2) is able to capture native speakers' intuitions about implicit causality in a sentence completion task. We first reproduce earlier results (showing lower surprisal values for pronouns that are congruent with either the subject or object, depending on which one corresponds to the implicit causality bias of the verb), and then examine the effects of gender and verb frequency on model performance. Our second study examines the reasoning ability of GPT-2: is the model able to produce more sensible motivations for why the subject VERBed the object if the verbs have stronger causality biases? We also developed a methodology to avoid human raters being biased by obscenities and disfluencies generated by the model.