ClinDet-Bench - Explanation Task: leaderboard
Metric: Share of the 16 clinical scoring systems (A-DROP, Alvarado, Apgar, BISAP, Bishop, Blatchford, CHADS2, Child-Pugh, CURB-65, Glasgow-Imrie, HAS-BLED, Khorana, Centor, ORBIT, qSOFA, SIRS) whose components and scoring rules the model explains correctly when asked one-shot, judged by a board-certified physician (proportion, 0-1); temperature 1.0, other settings default; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 0.94 | #64 |
| 2 | GPT-4o (2024-11-20) | 0.94 | #369 |
| 3 | Claude Opus 4.5 (20251101) | 0.94 | #70 |
| 4 | GPT-5.2 | 0.88 | #105 |
| 5 | DeepSeek V3.2 | 0.88 | #198 |
| 6 | DeepSeek R1 0528 | 0.81 | #217 |
| 7 | Llama 4 Maverick | 0.69 | #451 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=clindet-bench-explanation-task · How It Works · Data refreshed daily, snapshot 2026-10-11.