LLMs fail to replicate cross-cultural depression symptom differences

A new study published on arXiv investigated whether Large Language Models (LLMs) can accurately reflect cultural differences in depression symptom expression. Researchers found that LLMs, when prompted in English, largely failed to replicate the observed patterns where Western individuals report psychological symptoms and Eastern individuals report somatic ones. However, prompting the models in major Eastern languages like Chinese, Japanese, and Hindi showed some improvement in alignment. The study identified a lack of sensitivity to cultural personas and a strong, culturally invariant symptom hierarchy as key reasons for the LLMs' failure, highlighting that current general-purpose LLMs are not yet robust enough for safe and effective mental health applications. AI

IMPACT Current LLMs lack the cultural nuance required for safe and effective mental health applications, indicating a need for further development in cross-cultural understanding.

RANK_REASON The cluster contains an academic paper detailing research findings on LLMs. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• arXiv
• Large Language Models
• LLMs
• Shintaro Sakai
• Somatic in the East, Psychological in the West?: Investigating Clinically-Grounded Cross-Cultural Depression Symptom Expression in LLMs

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →