Potential and Limitations of LLMs in Capturing Structured Semantics: A Case Study on SRL
This addresses the controversy over LLMs' ability to understand structured semantics, which is crucial for improving language understanding and interpretability, though it is incremental in assessing existing models.
The study investigated whether Large Language Models (LLMs) can capture structured semantics using Semantic Role Labeling (SRL) as a test case, finding that LLMs show potential in this area but have limitations, with about 30% of errors overlapping with those made by untrained humans.
Large Language Models (LLMs) play a crucial role in capturing structured semantics to enhance language understanding, improve interpretability, and reduce bias. Nevertheless, an ongoing controversy exists over the extent to which LLMs can grasp structured semantics. To assess this, we propose using Semantic Role Labeling (SRL) as a fundamental task to explore LLMs' ability to extract structured semantics. In our assessment, we employ the prompting approach, which leads to the creation of our few-shot SRL parser, called PromptSRL. PromptSRL enables LLMs to map natural languages to explicit semantic structures, which provides an interpretable window into the properties of LLMs. We find interesting potential: LLMs can indeed capture semantic structures, and scaling-up doesn't always mirror potential. Additionally, limitations of LLMs are observed in C-arguments, etc. Lastly, we are surprised to discover that significant overlap in the errors is made by both LLMs and untrained humans, accounting for almost 30% of all errors.