MM-JudgeBias - Image Dominance: leaderboard
MM-JudgeBias (2026) tests multimodal LLM judges for compositional bias: the judge scores image, query and response triplets from 1 to 10 before and after a controlled perturbation. Bias-Deviation from 0 to 1, the normalized score drop when the query is emptied; higher is better.
Metric: Bias-Deviation (0 to 1) on MM-JudgeBias (about 200 image-query-response triplets per bias type, 1,804 in all, each judged on a 1-10 scale before and after a controlled perturbation; mean of three runs): mean drop of the judge's score, normalized by the largest possible drop, when the query is replaced with empty text; higher is better. Source: arxiv.org. Saturation forecast: Around September 2027. 26 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 4.5 | Thinking | 0.589 | #165 |
| 2 | Claude Sonnet 4.5 | Thinking | 0.54 | #267 |
| 3 | Gemini 2.5 Pro | 0.535 | #280 | |
| 4 | Claude Haiku 4.5 | Thinking | 0.53 | #587 |
| 5 | O4 Mini | High | 0.446 | #372 |
| 6 | Gemini 2.5 Flash Lite | 0.367 | #972 | |
| 7 | Gemini 2.0 Flash Lite | 0.358 | #1064 | |
| 8 | O3 | High | 0.354 | #242 |
| 9 | Qwen 3 VL 30B A3B (Thinking) | 0.343 | #727 | |
| 10 | Prometheus-Vision-13B | 0.34 | ||
| 11 | Gemini 2.5 Flash | 0.292 | #466 | |
| 12 | Gemini 3 Pro | High | 0.278 | #146 |
| 13 | Qwen 3 VL 8B Instruct | 0.266 | #1043 | |
| 14 | LLaVA-Critic-7B | 0.266 | ||
| 15 | InternVL3.5-14B | 0.243 | #1519 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=mm-judgebias-image-dominance · How It Works · Data refreshed daily, snapshot 2026-11-22.