SurgiQ - Negative: leaderboard
Negatively phrased all-except questions of SurgiQ (2026), a four-option surgical multiple-choice benchmark: accuracy in percent, reading the option label with the highest next-token probability, zero-shot. Higher is better.
Metric: Negative (all-except) questions (2,153): accuracy (%) on SurgiQ's four-option surgical multiple-choice questions, zero-shot; the answer is the option label with the highest next-token log-probability (plain and space-prefixed labels), options shuffled once. Source: arxiv.org. Saturation forecast: Around February 2027. 35 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 72B | 68.31 | #1190 |
| 2 | Qwen 2.5 72B Instruct | 68.03 | #1092 |
| 3 | Qwen 2.5 32B Instruct | 67.01 | #1215 |
| 4 | Ministral-3-14B-Reasoning-2512 | 64.96 | #1396 |
| 5 | Qwen 2.5 32B | 64.82 | #1332 |
| 6 | DeepSeek R1 Distill Qwen 32B | 64.36 | #1788 |
| 7 | GPT-OSS-120B | 62.08 | #700 |
| 8 | Ministral-3-8B-Reasoning-2512 | 59.48 | #1983 |
| 9 | DeepSeek R1 Distill Qwen 14B | 59.25 | #2582 |
| 10 | GPT-OSS-20B | 48.98 | #1147 |
| 11 | Gemma 2 9B (IT) | 46.1 | #2262 |
| 12 | Ministral-3-8B-Instruct-2512 | 43.12 | #1858 |
| 13 | DeepSeek-R1-Distill-Qwen-7B | 34.15 | #7996 |
| 14 | Gemma 2 2B (IT) | 33.09 | #6876 |
| 15 | MedGemma-4B-IT | 32.39 | #2716 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=surgiq-negative · How It Works · Data refreshed daily, snapshot 2026-11-13.