MMOOC - Misleading Premise (Yes/No): leaderboard

Metric: Shifted in-context score (%; mean of accuracy on the answerable Misleading Premise Yes/No questions and answer rationality, the mean 0-1 judge score of the answer's correctness and evidence consistency 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 December 2026. 18 models tracked.

Top models

#ModelScore
1Claude Opus 4.688.25
2Qwen 3.5 27B88.25
3GPT-4o82.5
4Gemma 4 26B81.75
5Qwen 3.5 122B A10B80.25
6O179.25
7O378.25
8Llama 4 Maverick76
9Gemma 4 31B75.75
10Ministral 3 14B75.25
11Ministral 3 8B73.25
12InternVL3-8B71.75
13Gemini 3.1 Pro (Preview)70.75

Interactive version: theaggregate.ai/benchmark?slug=mmooc-misleading-premise-yes-no · How It Works · Data refreshed daily, snapshot 2026-09-29.