Codeforces
A widely used online competitive programming platform that provides algorithmic contest problems. It is also employed as a benchmark for evaluating the algorithmic programming capabilities of AI models.
Timeline (last 90 days)
Model scores on competition mathematics (such as AIME and IMO) and code generation (such as Codeforces competition problems) have shown a near-vertical growth curve over the past year.
The vast majority of top competitors on competitive programming platforms such as Codeforces and AtCoder use C++.
AlphaCode was a milestone from DeepMind in 2022, reaching the median level of human competitors on Codeforces.
LiveCodeBench continuously scrapes new problems from competitive programming platforms such as LeetCode, Codeforces, and AtCoder to prevent data contamination.
All Facts (8)
o1模型在国际数学奥林匹克预选题上的表现接近金牌水平，在Codeforces编程竞赛中达到了89百分位
85%Unverified在Pass@1 Codeforces基准测试中，o1和o1 pro得分非常接近
80%UnverifiedCodeforces拥有超过200万注册用户
75%UnverifiedAlphaCode was a milestone from DeepMind in 2022, reaching the median level of human competitors on Codeforces.
60%UnverifiedIn the Codeforces rating, o1's score is far lower than that of o3-mini-high.
60%UnverifiedModel scores on competition mathematics (such as AIME and IMO) and code generation (such as Codeforces competition problems) have shown a near-vertical growth curve over the past year.
50%UnverifiedThe vast majority of top competitors on competitive programming platforms such as Codeforces and AtCoder use C++.
50%UnverifiedLiveCodeBench continuously scrapes new problems from competitive programming platforms such as LeetCode, Codeforces, and AtCoder to prevent data contamination.
50%