# Whose Assessment of Distress? Community Perspectives and LLM Alignment on Well-Being Posts

arXiv: https://arxiv.org/abs/2608.29446 · HTML: https://mlsanity.com/paper/2608.29446/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 66 references that are in the map.
- First posted on arXiv 29 Aug 2026.

## Builds on (most cited first, 19 of 19)

- [Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting](https://mlsanity.com/paper/2305.04388/) (2023)
- [Whose Opinions Do Language Models Reflect?](https://mlsanity.com/paper/2303.17548/) (2023)
- [Snorkel: Rapid Training Data Creation with Weak Supervision](https://mlsanity.com/paper/1711.10160/) (2017)
- [The Pushshift Reddit Dataset](https://mlsanity.com/paper/2001.08435/) (2020)
- [Annotators with Attitudes: How Annotator Beliefs And Identities Bias Toxic Language Detection](https://mlsanity.com/paper/2111.07997/) (2021)
- [Two Tales of Persona in LLMs: A Survey of Role-Playing and Personalization](https://mlsanity.com/paper/2406.01171/) (2024)
- [Fine-tuning language models to find agreement among humans with diverse preferences](https://mlsanity.com/paper/2211.15006/) (2022)
- [A Roadmap to Pluralistic Alignment](https://mlsanity.com/paper/2402.05070/) (2024)
- [Quantifying the Persona Effect in LLM Simulations](https://mlsanity.com/paper/2402.10811/) (2024)
- [Artificial Artificial Artificial Intelligence: Crowd Workers Widely Use Large Language Models for Text Production Tasks](https://mlsanity.com/paper/2306.07899/) (2023)
- [Sensitivity, Performance, Robustness: Deconstructing the Effect of Sociodemographic Prompting](https://mlsanity.com/paper/2309.07034/) (2023)
- [SMHD: A Large-Scale Resource for Exploring Online Language Usage for Multiple Mental Health Conditions](https://mlsanity.com/paper/1806.05258/) (2018)
- [Mental-LLM: Leveraging Large Language Models for Mental Health Prediction via Online Text Data](https://mlsanity.com/paper/2307.14385/) (2023)
- [DataDecide: How to Predict Best Pretraining Data with Small Experiments](https://mlsanity.com/paper/2504.11393/) (2025)
- [Sociodemographic Prompting is Not Yet an Effective Approach for Simulating Subjective Judgments with LLMs](https://mlsanity.com/paper/2311.09730/) (2023)
- [ComPO: Community Preferences for Language Model Personalization](https://mlsanity.com/paper/2410.16027/) (2024)
- [Simulating Psychological Risks in Human-AI Interactions: Real-Case Informed Modeling of AI-Induced Addiction, Anorexia, Depression, Homicide, Psychosis, and Suicide](https://mlsanity.com/paper/2511.08880/) (2025)
- [Cross-Cultural Differences in Mental Health Expressions on Social Media](https://mlsanity.com/paper/2402.11477/) (2024)
- [LOST: A Mental Health Dataset of Low Self-esteem in Reddit Posts](https://mlsanity.com/paper/2306.05596/) (2023)

## Abstract (arXiv)

Judgments about psychological distress are socially situated: what counts as concerning hinges on community norms around emotional expression, vulnerability, and help-seeking. Yet large language models (LLMs) used for distress detection are typically aligned to a single, undifferentiated standard. How well do these models capture the perspectives of the communities whose language they assess? We address this question through a perspectivist annotation study in which 321 participants provided 9,587 judgments on 1,198 Reddit posts spanning six identity-based communities, yielding community-specific labels. Raters in the contextualized in-group condition show a modest tendency to agree more with their community than uncontextualized out-group raters (OR = 1.18), an effect varying significantly across communities. We then evaluate nine open-weight LLM configurations and four frontier configurations against these labels. Open-weight LLMs systematically over-estimate distress: when communities perceive none-to-mild distress, these models achieve only 31-44% accuracy, predominantly producing false positives. GPT-5 and Gemini 2.5 Pro show the same none-to-mild inflation even when their full-sample over/under rates are mixed, while Claude Opus 4 is more conservative. This pattern does not simply mirror an outsider reading position: uncontextualized out-group human aggregates were nearly symmetric, with 18% over-estimation versus 19% under-estimation. Instead, the models that inflate none-to-mild cases exhibit a distress prior that exceeds both contextualized in-group and uncontextualized out-group human judgments. These findings have implications for equitable AI deployment in mental health contexts, where miscalibrated distress detection may unevenly affect the communities being assessed.