Modality Arbitration - MC2 - CoT Image-Target Accuracy: leaderboard
Modality Arbitration v1 Table3, MC2, Image-P image-groundtruth accuracy. MethodCoT; BF16,batch1,seed42,greedy,128tokens Base/512CoT. CorrelationQA uses added visual-reference annotation. Cross-target errors, steering methods and blur interventions are not substituted.
Metric: Score (%). Source: arxiv.org. Saturation forecast: Around February 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 VL 7B Instruct | 49.25 | #1842 |
| 2 | Qwen 3 VL 8B Instruct | 41.65 | #982 |
| 3 | InternVL3-8B | 37.4 | #1695 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=modality-arbitration-mc2-cot-image-target-accuracy · How It Works · Data refreshed daily, snapshot 2026-11-12.