Search papers, labs, and topics across Lattice.
This paper analyzes LLM reasoning processes to identify and categorize stigmatizing language towards individuals with mental health conditions, revealing biases missed by traditional multiple-choice evaluations. They develop a framework, informed by clinical expertise, to tag and rate the severity of stigmatizing statements within LLM reasoning steps. Results show that analyzing reasoning exposes significantly more stigma and highlights flaws in LLM logic regarding mental health than MCQ-based methods.
LLMs harbor surprisingly nuanced and pervasive mental health stigma, revealed only by dissecting their reasoning steps, not just their final answers.
While large language models (LLMs) are increasingly being explored for mental health applications, recent studies reveal that they can exhibit stigma toward individuals with psychological conditions. Existing evaluations of this stigma primarily rely on multiple-choice questions (MCQs), which fail to capture the biases embedded within the models'underlying logic. In this paper, we analyze the intermediate reasoning steps of LLMs to uncover hidden stigmatizing language and the internal rationales driving it. We leverage clinical expertise to categorize common patterns of stigmatizing language directed at individuals with psychological conditions and use this framework to identify and tag problematic statements in LLM reasoning. Furthermore, we rate the severity of these statements, distinguishing between overt prejudice and more subtle, less immediately harmful biases. To broaden the reasoning domain and capture a wider array of patterns, we also extend an existing mental health stigma benchmark by incorporating additional psychological conditions. Our findings demonstrate that evaluating model reasoning not only exposes substantially more stigma than traditional MCQ-based methods but it helps to identify the flaws in the LLMs'logic and their understanding of mental health conditions.
