Evidence / original sources

A claim should lead somewhere.

Follow the references behind Medical Evals AI. Each entry explains its role, links to the original publication, and identifies the guides that use it.

Content updated September 28, 2026. Inclusion does not imply endorsement.

01

Zuo et al. / Tsinghua University and Shanghai AI Laboratory

MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding ↗

Versioned construction, options, zero-shot evaluation and selected result tables.

Version: arXiv v3, 2025-06-06 / ICML 2025

Evidence locator: §3.1–3.2, §4.1; Tables 4–5

arxiv.org

Used inMedXpertQA Text and MM dossier →MedXpertQA Text versus MM: a higher score does not isolate an image benefit →MedXpertQA: difficulty filtering and public-question exposure →Read MedXpertQA results with the sample flag still attached →

02

TsinghuaC3I

MedXpertQA official repository ↗

Official release links and evaluation implementation.

Version: Accessed 2026-09-28

Evidence locator: README and eval

github.com

Used inMedXpertQA Text and MM dossier →MedXpertQA Text versus MM: a higher score does not isolate an image benefit →MedXpertQA: difficulty filtering and public-question exposure →Read MedXpertQA results with the sample flag still attached →

03

TsinghuaC3I

MedXpertQA official dataset card ↗

Separate Text/MM dev and test configurations; MIT license declaration.

Version: Accessed 2026-09-28

Evidence locator: Dataset card configuration

huggingface.co

Used inMedXpertQA Text and MM dossier →MedXpertQA Text versus MM: a higher score does not isolate an image benefit →MedXpertQA: difficulty filtering and public-question exposure →Read MedXpertQA results with the sample flag still attached →