Healthcare LLM Benchmarks Are Only as Good as Their Explicit Assumptions
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910246390202368 |
|---|---|
| author | Raman, Naveen Cortes-Gomez, Santiago Rubio, Mateo Dulce Fang, Fei Wilder, Bryan |
| author_facet | Raman, Naveen Cortes-Gomez, Santiago Rubio, Mateo Dulce Fang, Fei Wilder, Bryan |
| contents | Benchmarks are necessary for healthcare evaluation, but are not sufficient for predicting deployment performance. Our position is that the evaluation--deployment gap arises not because of poorly designed benchmarks, but from implicit assumptions about how users interact with models that cannot be surfaced from benchmarks alone. To make this precise, we propose a classification of assumptions into two categories: task, which can be tested from conversation data alone, and outcome, which requires outcome data and behavioral studies for testing. Critically, outcome assumptions depend on human behavior, something that even well-designed benchmarks cannot directly observe. To demonstrate the operationality of this framework, we retrospectively analyze a healthcare RCT as a case study and find that the gap naturally separates into task and outcome gaps of roughly equal size. To address this, we make two contributions: first, we propose BenchmarkCards, an artifact that documents assumptions, and second, we propose staged evaluation, a procedure that systematically tests assumptions and evaluates performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_22612 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Healthcare LLM Benchmarks Are Only as Good as Their Explicit Assumptions Raman, Naveen Cortes-Gomez, Santiago Rubio, Mateo Dulce Fang, Fei Wilder, Bryan Computers and Society Artificial Intelligence Machine Learning Benchmarks are necessary for healthcare evaluation, but are not sufficient for predicting deployment performance. Our position is that the evaluation--deployment gap arises not because of poorly designed benchmarks, but from implicit assumptions about how users interact with models that cannot be surfaced from benchmarks alone. To make this precise, we propose a classification of assumptions into two categories: task, which can be tested from conversation data alone, and outcome, which requires outcome data and behavioral studies for testing. Critically, outcome assumptions depend on human behavior, something that even well-designed benchmarks cannot directly observe. To demonstrate the operationality of this framework, we retrospectively analyze a healthcare RCT as a case study and find that the gap naturally separates into task and outcome gaps of roughly equal size. To address this, we make two contributions: first, we propose BenchmarkCards, an artifact that documents assumptions, and second, we propose staged evaluation, a procedure that systematically tests assumptions and evaluates performance. |
| title | Healthcare LLM Benchmarks Are Only as Good as Their Explicit Assumptions |
| topic | Computers and Society Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2605.22612 |