Language Shapes Mental Health Evaluations in Large Language Models

• URL: http://arxiv.org/abs/2603.06910v1
• Date: Fri, 06 Mar 2026 22:15:09 GMT
• Title: Language Shapes Mental Health Evaluations in Large Language Models
• Abstract summary: This study investigates whether large language models (LLMs) exhibit cross-linguistic differences in mental health evaluations.<n>We examine two widely used models, GPT-4o and Qwen3, to assess whether prompt language systematically shifts mental health-related evaluations.
• Score: 4.845028079612543
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: This study investigates whether large language models (LLMs) exhibit cross-linguistic differences in mental health evaluations. Focusing on Chinese and English, we examine two widely used models, GPT-4o and Qwen3, to assess whether prompt language systematically shifts mental health-related evaluations and downstream decision outcomes. First, we assess models' evaluative orientation toward mental health stigma using multiple validated measurement scales capturing social stigma, self-stigma, and professional stigma. Across all measures, both models produce higher stigma-related responses when prompted in Chinese than in English. Second, we examine whether these differences also manifest in two common downstream decision tasks in mental health. In a binary mental health stigma detection task, sensitivity to stigmatizing content varies across language prompts, with lower sensitivity observed under Chinese prompts. In a depression severity classification task, predicted severity also differs by prompt language, with Chinese prompts associated with more underestimation errors, indicating a systematic downward shift in predicted severity relative to English prompts. Together, these findings suggest that language context can systematically shape evaluative patterns in LLM outputs and shift decision thresholds in downstream tasks.

Related papers

• Creating Multilingual Mental Health Dialogue Datasets: Limits of Persona-Based Localization via Nationality and Language [11.022811270374675]
  AI and large language models (LLMs) have emerged as promising tools to address global mental health challenges.<n>Despite the global nature of these challenges, there remains a critical shortage of high-quality datasets for training and evaluating such systems.<n>This paper investigates whether similar persona-based methods can be used to generate multilingual mental health datasets.
  arXiv  Detail & Related papers  (2026-06-17T22:36:06Z)
• When Meanings Meet: Investigating the Emergence and Quality of Shared Concept Spaces during Multilingual Language Model Training [57.230355403478995]
  We investigate the development of language-agnostic concept spaces during pretraining of EuroLLM.<n>We find that shared concept spaces emerge early and continue to refine, but that alignment with them is language-dependent.<n>In contrast to prior work, our fine-grained manual analysis reveals that some apparent gains in translation quality reflect shifts in behavior.
  arXiv  Detail & Related papers  (2026-01-30T11:23:01Z)
• A Comprehensive Evaluation of Multilingual Chain-of-Thought Reasoning: Performance, Consistency, and Faithfulness Across Languages [48.68444770923683]
  We present the first comprehensive study of multilingual Chain-of-Thought (CoT) reasoning.<n>We measure language compliance, answer accuracy, and answer consistency when LRMs are prompt-hacked to think in a target language.<n>We find that the quality and effectiveness of thinking traces vary substantially depending on the prompt language.
  arXiv  Detail & Related papers  (2025-10-10T17:06:50Z)
• From Monolingual to Bilingual: Investigating Language Conditioning in Large Language Models for Psycholinguistic Tasks [9.837135712999471]
  Large Language Models (LLMs) exhibit strong linguistic capabilities, but little is known about how they encode psycholinguistic knowledge across languages.<n>We investigate whether and how LLMs exhibit human-like psycholinguistic responses under different linguistic identities.<n>We evaluate two models, Llama-3.3-70B-Instruct and Qwen2.5-72B-Instruct, under monolingual and bilingual prompting in English, Dutch, and Chinese.
  arXiv  Detail & Related papers  (2025-08-04T15:10:44Z)
• Mechanistic Understanding and Mitigation of Language Confusion in English-Centric Large Language Models [56.61984030508691]
  We present the first mechanistic interpretability study of language confusion.<n>We show that confusion points (CPs) are central to this phenomenon.<n>We show that editing a small set of critical neurons, identified via comparative analysis with a multilingual-tuned counterpart, substantially mitigates confusion.
  arXiv  Detail & Related papers  (2025-05-22T11:29:17Z)
• Enhancing Depression Detection with Chain-of-Thought Prompting: From Emotion to Reasoning Using Large Language Models [9.43184936918456]
  Depression is one of the leading causes of disability worldwide.<n>Recent advancements in Large Language Models have shown promise in addressing mental health challenges.<n>We propose a Chain-of-Thought Prompting approach that enhances both the performance and interpretability of depression detection.
  arXiv  Detail & Related papers  (2025-02-09T12:30:57Z)
• LlaMADRS: Prompting Large Language Models for Interview-Based Depression Assessment [75.44934940580112]
  This study introduces LlaMADRS, a novel framework leveraging open-source Large Language Models (LLMs) to automate depression severity assessment.<n>We employ a zero-shot prompting strategy with carefully designed cues to guide the model in interpreting and scoring transcribed clinical interviews.<n>Our approach, tested on 236 real-world interviews, demonstrates strong correlations with clinician assessments.
  arXiv  Detail & Related papers  (2025-01-07T08:49:04Z)
• Large Language Models as Neurolinguistic Subjects: Discrepancy between Performance and Competence [49.60849499134362]
  This study investigates the linguistic understanding of Large Language Models (LLMs) regarding signifier (form) and signified (meaning)<n>We introduce a neurolinguistic approach, utilizing a novel method that combines minimal pair and diagnostic probing to analyze activation patterns across model layers.<n>We found: (1) Psycholinguistic and neurolinguistic methods reveal that language performance and competence are distinct; (2) Direct probability measurement may not accurately assess linguistic competence; and (3) Instruction tuning won't change much competence but improve performance.
  arXiv  Detail & Related papers  (2024-11-12T04:16:44Z)
• Building Multilingual Datasets for Predicting Mental Health Severity through LLMs: Prospects and Challenges [3.0382033111760585]
  Large Language Models (LLMs) are increasingly being integrated into various medical fields, including mental health support systems.<n>We present a novel multilingual adaptation of widely-used mental health datasets, translated from English into six languages.<n>This dataset enables a comprehensive evaluation of LLM performance in detecting mental health conditions and assessing their severity across multiple languages.
  arXiv  Detail & Related papers  (2024-09-25T22:14:34Z)
• Adapting Mental Health Prediction Tasks for Cross-lingual Learning via Meta-Training and In-context Learning with Large Language Model [3.3590922002216193]
  We use model-agnostic meta-learning and leveraging large language models (LLMs) to address this gap. We first apply a meta-learning model with self-supervision, which results in improved model initialisation for rapid adaptation and cross-lingual transfer. In parallel, we use LLMs' in-context learning capabilities to assess their performance accuracy across the Swahili mental health prediction tasks.
  arXiv  Detail & Related papers  (2024-04-13T17:11:35Z)
• Semantic Coherence Markers for the Early Diagnosis of the Alzheimer Disease [0.0]
  Perplexity was originally conceived as an information-theoretic measure to assess how much a given language model is suited to predict a text sequence. We employed language models as diverse as N-grams, from 2-grams to 5-grams, and GPT-2, a transformer-based language model. Best performing models achieved full accuracy and F-score (1.00 in both precision/specificity and recall/sensitivity) in categorizing subjects from both the AD class and control subjects.
  arXiv  Detail & Related papers  (2023-02-02T11:40:16Z)
• AM2iCo: Evaluating Word Meaning in Context across Low-ResourceLanguages with Adversarial Examples [51.048234591165155]
  We present AM2iCo, Adversarial and Multilingual Meaning in Context. It aims to faithfully assess the ability of state-of-the-art (SotA) representation models to understand the identity of word meaning in cross-lingual contexts. Results reveal that current SotA pretrained encoders substantially lag behind human performance.
  arXiv  Detail & Related papers  (2021-04-17T20:23:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.