Ask HN: Claude Opus 的性能受一天中时间的影响吗？
10 分•作者: scaredreally•6 个月前
我非常喜欢 Claude Opus，因为它非常擅长理解功能请求，并且通常能与我的代码库（完全使用 Opus 从头编写）保持一致。<p>我最近注意到，当我在晚上（美国东部时间）使用 Opus 时，它会在我经常提出的相同类型的请求上陷入极端的“兔子洞”。它更有可能进行破坏代码的重构，然后在这些错误上进行迭代，形成一种螺旋式循环。一个通常需要 3-4 分钟的请求会变成一个 10 分钟的冒险，直到我回滚更改，指出错误，然后重试。它会很乐意承认错误，但这种模式似乎很一致。<p>我还没有进行过类似的测试，这会很有趣，但其他人也注意到同样的情况了吗？
查看原文
I am a big fan of Claude Opus as it has been very good at understanding feature requests and generally staying consistent with my codebase (completely written from scratch using Opus).<p>I've noticed recently that when I am using Opus at night (Eastern US), I am seeing it go down extreme rabbit holes on the same types of requests I am putting through on a regular basis. It is more likely to undertake refactors that break the code and then iterates on those errors in a sort of spiral. A request that would normally take 3-4 minutes will turn into a 10 minute adventure before I revert the changes, call out the mistake, and try again. It will happily admit the mistake, but the pattern seems to be consistent.<p>I haven't performed a like for like test and that would be interesting, but has anyone else noticed the same?