HealthBench Professional GuideEvidence / original sources
Follow the references behind HealthBench Professional Guide. 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.
OpenAI
Pinned reference scoring implementation, including positive-point normalization.
Version: Commit 652c89d0ca9df547706735883097e9537d40dc47
Evidence locator: calculate_score; calculate_length_adjusted_score; aggregate results
github.comRebecca Soskin Hicks and colleagues · OpenAI
Primary sampling, length adjustment, model/harness comparison and composition tables.
Version: arXiv v1 · 2026-04-30
Evidence locator: Sections 3–6 and 8; Appendix Tables 1–3
arxiv.orgOpenAI
MIT metadata, field definitions, contamination request and internal-evaluator limitation.
Version: Accessed 2026-09-28
Evidence locator: Dataset card; implementation note
huggingface.coOpenAI
Dated announcement introducing the benchmark; separate from the subsequent arXiv version.
Version: Published 2026-04-22
Evidence locator: Continuing to evaluate and strengthen model health performance and safety
openai.com