Coding Benchmarks Are Dying. MirrorCode Asks AI to Rebuild the Whole Program.
中文摘要
传统代码基准测试已过时，MirrorCode 提出让 AI 重构整个程序的新方法，以更好地评估前沿智能体的能力。
English Summary
Traditional coding benchmarks are obsolete; MirrorCode proposes having AI rebuild entire programs to better evaluate the capabilities of frontier agents.
原文节选
Programming tests used to measure whether AI could solve a difficult coding problem. Now frontier agents can work for days, spend billions… Continue reading on Medium »