Will Gemini achieve a score above 90% on the MMLU benchmark?
21
984
410
resolved Dec 6
Resolved
YES

Considering the performance improvement resulting from progress of prompt engineering, I will have a deadline of 2025/1/1.

Get Ṁ200 play money

🏅 Top traders

#NameTotal profit
1Ṁ126
2Ṁ78
3Ṁ31
4Ṁ21
5Ṁ14
Sort by:

The exact score is 90.04% - This Question will resolve YES.

@3684 Curious where you found that

Does it have to be one shot?

@glassbottle No. All types of prompts are allowed.

lol

pure LLM or anything using Gemini as a submodel?

@NikhilVyas pure LLM

More related questions