MMOOC - Uncertain Spatial and Physical Context (Yes/No): leaderboard

Metric: Out-of-context score (%; mean of the refusal rate, the share of the Uncertain Spatial and Physical Context Yes/No questions, which the image and context cannot answer, that the model correctly declines, and refusal rationality, the mean 0-1 judge score of its abstention reasoning in steps of 0.25; judge scores averaged over GPT-5.6, Claude Opus 5 and DeepSeek-V4-Pro). Source: arxiv.org. Saturation forecast: Around 2036. 18 models tracked.

Top models

#ModelScore
1O144.75
2Gemma 4 31B35.25
3GPT-4o30
4Ministral 3 14B28.5
5Gemma 4 26B27.75
6Ministral 3 8B26.5
7Qwen 3.5 122B A10B26.25
8Qwen 3.5 27B22.5
9Gemini 3.1 Pro (Preview)21.75
10Llama 4 Maverick19
11O318
12Claude Opus 4.615.25
13InternVL3-8B12

Interactive version: theaggregate.ai/benchmark?slug=mmooc-uncertain-spatial-and-physical-context-yes-no · How It Works · Data refreshed daily, snapshot 2026-09-29.