ScholScan (Image Input) - Data Handling and Preprocessing: leaderboard
Metric: Flaw-detection score (0-1, scaled to 0-100): per question, zero unless the model finds the annotated scientific error, otherwise the geometric mean of evidence-location Dice and reasoning-chain prefix match, times a penalty for unrelated errors listed, as extracted by a GPT-4.1 evaluator, averaged over the data handling and preprocessing questions of ScholScan's 1,800 expert-reviewed questions that ask a model to scan a whole academic paper (ICLR 2024-2025 and Nature Communications, 715 papers) for a planted or reviewer-reported scientific flaw without being told where to look, with the paper given as page images; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 27 | #145 |
| 2 | GPT-5 | 26.6 | #91 |
| 3 | Doubao-Seed-1.6 (Thinking) | 15.1 | |
| 4 | Seed-1.6 | 15 | #257 |
| 5 | Grok 4 | 7.4 | #169 |
| 6 | Llama 4 Maverick | 4 | #451 |
| 7 | Gemma 3 27B | 1 | #596 |
| 8 | Mistral Small 3.1 | 0.1 | #600 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=scholscan-image-input-data-handling-and-preprocessing · How It Works · Data refreshed daily, snapshot 2026-10-11.