What will the top score on Humanity's Last Exam be when it is released?
Plus
10
Ṁ1054Feb 9
74%
>10%
53%
>20%
27%
>30%
19%
>40%
9%
>50%
https://www.safe.ai/blog/humanitys-last-exam
Humanity's Last Exam is a new benchmark developed by CAIS and Scale AI to measure AI's proximity to expert-level capabilities. The goal is to create the world's most difficult AI test by gathering complex questions from experts across fields. The project invites experts to submit challenging questions, with accepted contributors offered co-authorship on the resulting paper and prizes from a $500,000 pool.
Currently, to qualify for the benchmark, all models must fail the submitted question.
This question is managed and resolved by Manifold.
Get
1,000
and3.00
Related questions
Related questions
What will be the best normalized score achieved on the original 7 RE-Bench tasks by December 31st 2025?
What will be the best score on Cybench by December 31st 2025?
Will an AI score over 80% on FrontierMath Benchmark in 2025
20% chance
Will there be another benchmark/test after "Humanity's Last Exam"?
66% chance
Will any model get above human level on the Simple Bench benchmark before September 1st, 2025.
69% chance
Will an AI score over 30% on FrontierMath Benchmark in 2025
86% chance
What will be the best score on the WebArena benchmark before 2025?
64% chance
What will be the best score on the GAIA benchmark before 2025?
64% chance
What will be the best score on the SWE-Bench (unassisted) benchmark before 2025?
54% chance
When will SOTA for Atari 100k pass human median and mean score on all 57 games?