A Layered Multi-Expert Framework for Long-Context Mental Health Assessments
This addresses reliability issues in AI-driven mental health screening, though it appears incremental as it builds on existing multi-model approaches.
The paper tackled the problem of hallucinations and inconsistent reasoning in large language models when processing long-form mental health assessments, introducing a layered multi-expert framework that improved accuracy, F1-score, and reduced PHQ-8 error on datasets like DAIC-WOZ and curated case studies.
Long-form mental health assessments pose unique challenges for large language models (LLMs), which often exhibit hallucinations or inconsistent reasoning when handling extended, domain-specific contexts. We introduce Stacked Multi-Model Reasoning (SMMR), a layered framework that leverages multiple LLMs and specialized smaller models as coequal 'experts'. Early layers isolate short, discrete subtasks, while later layers integrate and refine these partial outputs through more advanced long-context models. We evaluate SMMR on the DAIC-WOZ depression-screening dataset and 48 curated case studies with psychiatric diagnoses, demonstrating consistent improvements over single-model baselines in terms of accuracy, F1-score, and PHQ-8 error reduction. By harnessing diverse 'second opinions', SMMR mitigates hallucinations, captures subtle clinical nuances, and enhances reliability in high-stakes mental health assessments. Our findings underscore the value of multi-expert frameworks for more trustworthy AI-driven screening.