RoboAbstention - False Premise: leaderboard

The False Premise category of RoboAbstention (2026, The Yes-Man Syndrome): 957 instructions built on a false premise about the scene, given with an image to a vision-language robot planner that should abstain or ask. The score is the abstention rate in percent; higher is better.

Metric: Abstention rate (%) on the 957 False Premise instructions, image plus instruction to an embodied VLM planner, every instruction warranting abstention, act or abstain judged by GPT-5.4 mini (97.5 percent agreement with human labels), API default sampling; higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 8 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 2.5 Flash28.5#463
2Llama 4 Maverick15.5#1110
3Qwen 3.5 27B7.6#375
4Claude Sonnet 4.67.2#173
5Gemini Robotics-ER 1.67#170
6GPT-5.43.6#135
7GPT-5.4 MiniHigh3.4#406
8GPT-5.4 Nano1.5#651

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=roboabstention-false-premise · How It Works · Data refreshed daily, snapshot 2026-11-21.