Explaining Large Language Model-Based Neural Semantic Parsers (Student Abstract)

TL;DR

This paper investigates methods to explain how large language models perform neural semantic parsing, aiming to understand their success and guide future research in interpretability.

Contribution

It introduces explanation methods for LLM-based semantic parsers and provides qualitative analysis of their behaviors, addressing a gap in understanding their mechanisms.

Findings

Different explanation methods reveal diverse model behaviors

Qualitative analysis offers insights into model decision processes

Study aims to inspire further research in model interpretability

Abstract

While large language models (LLMs) have demonstrated strong capability in structured prediction tasks such as semantic parsing, few amounts of research have explored the underlying mechanisms of their success. Our work studies different methods for explaining an LLM-based semantic parser and qualitatively discusses the explained model behaviors, hoping to inspire future research toward better understanding them.