This shows up in OpenAI's graphs on their announcement page. There is a peak performance datapoint in the graphs past which (to the right on the graph indicating more resources spent) peformance declines. And it's on every graph on that page!
replyAnd in my tests, that point of "overthinking" depends on the problem's complexity, so it's not necessarily that using "xhigh" is always bad or good.
replyI think I have this problem but with my human brain.
reply