Predicting Sentence Acceptability Judgments in Multimodal Contexts

TL;DR

This study investigates how visual context influences sentence acceptability judgments in humans and large language models, revealing minimal impact on humans but notable effects on model predictions and internal representations.

Contribution

It demonstrates that visual images have little effect on human judgments but significantly influence LLM predictions and internal representations, highlighting differences in multimodal processing.

Findings

Humans' acceptability ratings are unaffected by visual context.

LLMs' predictions are slightly better without visual context.

Model judgments vary, with Qwen resembling human patterns.

Abstract

Previous work has examined the capacity of deep neural networks (DNNs), particularly transformers, to predict human sentence acceptability judgments, both independently of context, and in document contexts. We consider the effect of prior exposure to visual images (i.e., visual context) on these judgments for humans and large language models (LLMs). Our results suggest that, in contrast to textual context, visual images appear to have little if any impact on human acceptability ratings. However, LLMs display the compression effect seen in previous work on human judgments in document contexts. Different sorts of LLMs are able to predict human acceptability judgments to a high degree of accuracy, but in general, their performance is slightly better when visual contexts are removed. Moreover, the distribution of LLM judgments varies among models, with Qwen resembling human patterns, and others diverging from them. LLM-generated predictions on sentence acceptability are highly correlated with their normalised log probabilities in general. However, the correlations decrease when visual contexts are present, suggesting that a higher gap exists between the internal representations of LLMs and their generated predictions in the presence of visual contexts. Our experimental work suggests interesting points of similarity and of difference between human and LLM processing of sentences in multimodal contexts.