SciFigPlag-Bench - Source Attribution: leaderboard
Metric: Accuracy (%; pick which of four candidate figures the suspicious figure reuses, or none; SciFigPlag-Bench pairs of source and suspicious scientific figures (2,582 plagiarized pairs from documented real cases and taxonomy-guided synthesis, 2,541 visually similar negatives); the same task-specific prompts for every model, temperature 0). Source: arxiv.org. Saturation forecast: Estimated already saturated. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 98.1 |
| 2 | Gemini 3 Flash | 97.8 |
| 3 | Qwen 3.5 35B A3B | 97.8 |
| 4 | Qwen 3.6 35B A3B | 97.6 |
| 5 | GPT-5.4 | 97.1 |
| 6 | Gemma 4 31B (IT) | 92.6 |
| 7 | Pixtral-12B-2409 | 67.7 |
| 8 | InternVL3.5-8B | 64.4 |
| 9 | Gemma 4 26B A4B (IT) | 62.2 |
| 10 | Qwen 3.5 2B | 29.4 |
| 11 | Gemma 3 4B (IT) | 29.1 |
| 12 | Mistral Small 4 | 24.9 |
Interactive version: theaggregate.ai/benchmark?slug=scifigplag-bench-source-attribution · How It Works · Data refreshed daily, snapshot 2026-09-29.