Does Rationale Quality Matter? Enhancing Mental Disorder Detection via Selective Reasoning Distillation

2025-05-26

AI Summary

Reasoning distillation for mental health assessment is hindered by low-quality, domain-mismatched explanations generated by large language models (LLMs). Method: We propose a Clinical Reasoning Alignment framework featuring rationale filtering, integrating clinical symptom ontology modeling, multi-dimensional rationale relevance scoring, and alignment evaluation to construct a supervised reasoning distillation pipeline. Contribution/Results: First, we systematically validate the critical impact of rationale quality on distillation performance. Second, we design a clinical-knowledge-guided rationale selection mechanism that overcomes noise and domain misalignment inherent in end-to-end distillation. Evaluated across multiple mental disorder detection datasets, our approach improves small-model F1-score by 4.2% and explanation faithfulness by 19.7% over standard distillation baselines—demonstrating significant gains in both predictive accuracy and interpretability.

Technology Category

• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

The detection of mental health problems from social media and the interpretation of these results have been extensively explored. Research has shown that incorporating clinical symptom information into a model enhances domain expertise, improving its detection and interpretation performance. While large language models (LLMs) are shown to be effective for generating explanatory rationales in mental health detection, their substantially large parameter size and high computational cost limit their practicality. Reasoning distillation transfers this ability to smaller language models (SLMs), but inconsistencies in the relevance and domain alignment of LLM-generated rationales pose a challenge. This paper investigates how rationale quality impacts SLM performance in mental health detection and explanation generation. We hypothesize that ensuring high-quality and domain-relevant rationales enhances the distillation. To this end, we propose a framework that selects rationales based on their alignment with expert clinical reasoning. Experiments show that our quality-focused approach significantly enhances SLM performance in both mental disorder detection and rationale generation. This work highlights the importance of rationale quality and offers an insightful framework for knowledge transfer in mental health applications.

Problem

• Enhancing mental disorder detection via selective reasoning distillation
• Investigating rationale quality impact on smaller language models
• Proposing a framework for domain-aligned rationale selection

Innovation

• Selects rationales aligned with clinical reasoning
• Distills LLM reasoning into smaller models
• Enhances mental disorder detection performance