A Dual-Prompting for Interpretable Mental Health Language Models

TL;DR

This paper introduces a dual-prompting method that enhances the interpretability of mental health language models by extracting and summarizing evidence of suicidality, aiding clinicians in understanding model decisions.

Contribution

It presents a novel dual-prompting approach combining knowledge-aware evidence extraction and summarization to improve interpretability of mental health LLMs.

Findings

Improved performance in evidence extraction and summarization tasks.

Enhanced interpretability of mental health language models.

Potential to assist clinicians in mental health assessment.

Abstract

Despite the increasing demand for AI-based mental health monitoring tools, their practical utility for clinicians is limited by the lack of interpretability.The CLPsych 2024 Shared Task (Chim et al., 2024) aims to enhance the interpretability of Large Language Models (LLMs), particularly in mental health analysis, by providing evidence of suicidality through linguistic content. We propose a dual-prompting approach: (i) Knowledge-aware evidence extraction by leveraging the expert identity and a suicide dictionary with a mental health-specific LLM; and (ii) Evidence summarization by employing an LLM-based consistency evaluator. Comprehensive experiments demonstrate the effectiveness of combining domain-specific information, revealing performance improvements and the approach's potential to aid clinicians in assessing mental state progression.