Guardado en:
Detalles Bibliográficos
Autores principales: Gruber, Victoria-Elisabeth, Marinescu, Razvan, Fajardo, Diego, Nassar, Amin H., Arkfeld, Christopher, Ludlow, Alexandria, Patel, Shama, Samaei, Mehrnoosh, Klug, Valerie, Huber, Anna, Gühner, Marcel, Orfila, Albert Botta i, Lagoja, Irene, Tarr, Kimya, Larson, Haleigh, Howard, Mary Beth
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2512.17028
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908721617043456
author Gruber, Victoria-Elisabeth
Marinescu, Razvan
Fajardo, Diego
Nassar, Amin H.
Arkfeld, Christopher
Ludlow, Alexandria
Patel, Shama
Samaei, Mehrnoosh
Klug, Valerie
Huber, Anna
Gühner, Marcel
Orfila, Albert Botta i
Lagoja, Irene
Tarr, Kimya
Larson, Haleigh
Howard, Mary Beth
author_facet Gruber, Victoria-Elisabeth
Marinescu, Razvan
Fajardo, Diego
Nassar, Amin H.
Arkfeld, Christopher
Ludlow, Alexandria
Patel, Shama
Samaei, Mehrnoosh
Klug, Valerie
Huber, Anna
Gühner, Marcel
Orfila, Albert Botta i
Lagoja, Irene
Tarr, Kimya
Larson, Haleigh
Howard, Mary Beth
contents As large language models (LLMs) become primary sources of health information for millions, their accuracy in women's health remains critically unexamined. We introduce the Women's Health Benchmark (WHB), the first benchmark evaluating LLM performance specifically in women's health. Our benchmark comprises 96 rigorously validated model stumps covering five medical specialties (obstetrics and gynecology, emergency medicine, primary care, oncology, and neurology), three query types (patient query, clinician query, and evidence/policy query), and eight error types (dosage/medication errors, missing critical information, outdated guidelines/treatment recommendations, incorrect treatment advice, incorrect factual information, missing/incorrect differential diagnosis, missed urgency, and inappropriate recommendations). We evaluated 13 state-of-the-art LLMs and revealed alarming gaps: current models show approximately 60\% failure rates on the women's health benchmark, with performance varying dramatically across specialties and error types. Notably, models universally struggle with "missed urgency" indicators, while newer models like GPT-5 show significant improvements in avoiding inappropriate recommendations. Our findings underscore that AI chatbots are not yet fully able of providing reliable advice in women's health.
format Preprint
id arxiv_https___arxiv_org_abs_2512_17028
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Women's Health Benchmark for Large Language Models
Gruber, Victoria-Elisabeth
Marinescu, Razvan
Fajardo, Diego
Nassar, Amin H.
Arkfeld, Christopher
Ludlow, Alexandria
Patel, Shama
Samaei, Mehrnoosh
Klug, Valerie
Huber, Anna
Gühner, Marcel
Orfila, Albert Botta i
Lagoja, Irene
Tarr, Kimya
Larson, Haleigh
Howard, Mary Beth
Computation and Language
Artificial Intelligence
Machine Learning
As large language models (LLMs) become primary sources of health information for millions, their accuracy in women's health remains critically unexamined. We introduce the Women's Health Benchmark (WHB), the first benchmark evaluating LLM performance specifically in women's health. Our benchmark comprises 96 rigorously validated model stumps covering five medical specialties (obstetrics and gynecology, emergency medicine, primary care, oncology, and neurology), three query types (patient query, clinician query, and evidence/policy query), and eight error types (dosage/medication errors, missing critical information, outdated guidelines/treatment recommendations, incorrect treatment advice, incorrect factual information, missing/incorrect differential diagnosis, missed urgency, and inappropriate recommendations). We evaluated 13 state-of-the-art LLMs and revealed alarming gaps: current models show approximately 60\% failure rates on the women's health benchmark, with performance varying dramatically across specialties and error types. Notably, models universally struggle with "missed urgency" indicators, while newer models like GPT-5 show significant improvements in avoiding inappropriate recommendations. Our findings underscore that AI chatbots are not yet fully able of providing reliable advice in women's health.
title A Women's Health Benchmark for Large Language Models
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2512.17028