OncoTriad-QA - Tumor Grade: leaderboard
Metric: Accuracy (%) on the universal tumor-grade questions of OncoTriad-QA's held-out TCGA cases in the all-available modality setting (representative pathology ROI images, radiology slices, genomic summaries and clinical metadata): multiple choice, mean of three runs over a frozen stratified 60% sample of the held-out test items; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 3 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 52.2 |
| 2 | MedGemma-4B-IT | 40.3 |
Interactive version: theaggregate.ai/benchmark?slug=oncotriad-qa-tumor-grade · How It Works · Data refreshed daily, snapshot 2026-09-29.