Sens-VisualNews - Strict: leaderboard
Metric: Top-1 accuracy (%) on the test split of the strict subset (sensational images with unanimous annotator agreement and as many non-sensational images), zero-shot yes or no answer to whether the news image is sensational, greedy decoding, each model with the prompt and prompt format that scored best on the development split; balanced classes, chance 50; higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | LLaVA-OneVision-1.5-8B | 93.9 |
| 2 | Qwen3-VL-8B | 92.8 |
| 3 | Qwen3-VL-2B | 92.7 |
| 4 | InternVL3.5-8B | 92.1 |
| 5 | LLaVA-OneVision-7B | 91.8 |
| 6 | Qwen3-VL-4B | 90.8 |
| 7 | LLaVA-OneVision-0.5B | 90.8 |
| 8 | LLaVA-OneVision-1.5-4B | 89.7 |
| 9 | InternVL3.5-4B | 88.8 |
| 10 | InternVL3.5-2B | 86.2 |
Interactive version: theaggregate.ai/benchmark?slug=sens-visualnews-strict · How It Works · Data refreshed daily, snapshot 2026-10-07.