1 paper
Laura M. Vowels, Matthew J. Vowels, Shivali Sharma +9
People increasingly use large language models (LLMs) for mental health support, yet their safety in evolving, high-risk conversations remains poorly characterised. We developed K-B…