The Curious Case of Control

TL;DR

This paper investigates whether large language models exhibit systematic errors in subject control sentences similar to children and examines their sensitivity to semantic role cues, revealing distinct behavioral groups and the influence of prompting.

Contribution

It categorizes language models based on their control behavior, analyzes their reliance on positional heuristics, and explores how prompting affects their outputs and semantic role understanding.

Findings

Models fall into three behavioral groups with broad differences.

Largest group models rely on positional heuristics, succeeding on subject but not on object control.

Prompting with agent-patient information significantly alters model outputs.

Abstract

Children acquiring English make systematic errors on subject control sentences even after they have reached near-adult competence (C. Chomsky, 1969), possibly due to heuristics based on semantic roles (Maratsos, 1974). Given the advanced fluency of large generative language models, we ask whether model outputs are consistent with these heuristics, and to what degree different models are consistent with each other. We find that models can be categorized by behavior into three separate groups, with broad differences between the groups. The outputs of models in the largest group are consistent with positional heuristics that succeed on subject control but fail on object control. This result is surprising, given that object control is orders of magnitude more frequent in the text data used to train such models. We examine to what degree the models are sensitive to prompting with agent-patient information, finding that raising the salience of agent and patient relations results in significant changes in the outputs of most models. Based on this observation, we leverage an existing dataset of semantic proto-role annotations (White, et al. 2020) to explore the connections between control and labeling event participants with properties typically associated with agents and patients.