Project Euler as an AI benchmark Created 2025-03-24 Updated 2025-10-14
The beauty of Project Euler is that it would serve both as a AI code generation benchmark and as an AI Math benchmark!
Updates Getting banned from Project Euler 2025-10-27
I have been banned from Project Euler for life, and cannot login to my previous account projecteuler.net/profile/cirosantilli.png
The ban happened within 12 hours of me publishing a solution to Project Euler problem 961 github.com/lucky-bai/projecteuler-solutions/pull/94 which was one-shot by a free GPT-5 account as MathArena had alerted me to being possible: matharena.ai/?comp=euler--euler&task=4&model=GPT-5+%28high%29&run=1
The problem leaderboard contains several people solved the problem within minutes of it being released, so almost certainly with an LLM.
The "secret club" mentality is their only blemish, and incompatible with open science.
They should also make sure that LLMs don't one shot their future problems BEFORE publishing them!