Assessing consistency of AI chatbot responses in ophthalmology medical exams

Purpose: Large language models (LLMs) are increasingly evaluated in ophthalmology, often using single test iterations that overlook whether responses remain consistent under repeated conditions. We aim to assess commonly used AI models under multiple testing iterations with varying conditions, inclu...

Полное описание

Библиографические подробности

Главные авторы:	Elad Shvartz, Regina Raiyter, Amelia Goldshtein, Omri Zur, Eyal Margalit, Daniel Bahir
Формат:	Artigo
Язык:	Inglês
Опубликовано:	Elsevier 2025-12-01
Серии:	AJO International
Предметы:	ChatGPT Large language models (LLMs) Artificial intelligence (AI) Consistency Reliability GPT-4o
Online-ссылка:	http://www.sciencedirect.com/science/article/pii/S2950253525000930
Метки:	Добавить метку Нет меток, Требуется 1-ая метка записи!