[Schlagbohrer]

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!

[XCSme]

And 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.