Analyzing LLM Reasoning to Uncover Mental Health Stigma
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915962111918080 |
|---|---|
| author | Sankar, Sreehari Nafar, Aliakbar Barman, Mona Heitz, Hannah K. Kumar, Ashwin Tohidi, Pouria Li, Dailun Hussain, Danish DuBois, Russell Hasheminia, Hamed Majzoubi, Farshad |
| author_facet | Sankar, Sreehari Nafar, Aliakbar Barman, Mona Heitz, Hannah K. Kumar, Ashwin Tohidi, Pouria Li, Dailun Hussain, Danish DuBois, Russell Hasheminia, Hamed Majzoubi, Farshad |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_25053 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Analyzing LLM Reasoning to Uncover Mental Health Stigma Sankar, Sreehari Nafar, Aliakbar Barman, Mona Heitz, Hannah K. Kumar, Ashwin Tohidi, Pouria Li, Dailun Hussain, Danish DuBois, Russell Hasheminia, Hamed Majzoubi, Farshad Computation and Language Artificial Intelligence I.2.7 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. |
| title | Analyzing LLM Reasoning to Uncover Mental Health Stigma |
| topic | Computation and Language Artificial Intelligence I.2.7 |
| url | https://arxiv.org/abs/2604.25053 |