Researchers have conducted a comprehensive audit of benchmarks used for detecting depression from clinical interviews. Their analysis revealed significant discrepancies between different evaluation protocols, including cross-validation and official test splits, with top-ranked models often performing poorly when transferred to external datasets. The study also found that text-based models showed a notable increase in performance on symptom-dense interview segments compared to audio-based models. AI
IMPACT Highlights potential unreliability in AI models for mental health assessment, urging caution in deployment.
RANK_REASON The cluster contains an academic paper detailing a new evaluation methodology and findings. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
