ModernBert to LLaMA: a Benchmark Study of LLM Robustness for Depression Detection in Social Media Texts

Yash Prakash Narayan; Arun Kumar Dubey; Achin Jain; Arvind Panwar; Alex Varghese; Ondřej Krejcar; Lukáš Herout; Hamidreza Namazi

doi:10.1007/s44196-026-01636-x

SCHEDA SCIENTIFICA2026Altra ricerca

ModernBert to LLaMA: a Benchmark Study of LLM Robustness for Depression Detection in Social Media Texts

Yash Prakash Narayan, Arun Kumar Dubey, Achin Jain, Arvind Panwar, Alex Varghese, Ondřej Krejcar, Lukáš Herout, Hamidreza Namazi

International Journal of Computational Intelligence Systems · 2026

Psicopatologia Psicologia della salute

Abstract e pubblicazione originale

L’abstract è consultabile nella fonte originale. Non viene ripubblicato qui perché non risulta verificata una licenza compatibile.

Leggi nella fonte originale ↗

Parole chiave

large language modelsdepression detectionsocial media textrobustness benchmarkingBERTllamasmental health screeningtext classificationMental Health via WritingSentiment Analysis and Opinion MiningHate Speech and Cyberbullying Detection

Come citare questo studio

Yash Prakash Narayan, Arun Kumar Dubey, Achin Jain, Arvind Panwar, Alex Varghese, Ondřej Krejcar, Lukáš Herout, Hamidreza Namazi. (2026). ModernBert to LLaMA: a Benchmark Study of LLM Robustness for Depression Detection in Social Media Texts. International Journal of Computational Intelligence Systems. DOI: 10.1007/s44196-026-01636-x

Consulta la pubblicazione originale ↗