2 ms·
Not always, in some cases, changing to a higher reasoning makes the AI doubt itself too much, and skip over the correct answer by overcomplicating the problem a
by XCSme 3mo ago
Not always, in some cases, changing to a higher reasoning makes the AI doubt itself too much, and skip over the correct answer by overcomplicating the problem and polluting the context.
It would be nice to see on which categories of problems the extra thinking makes it better and on which it makes it worse.
- andai 3mo agoI think I have this problem but with my human brain.
- Schlagbohrer 3mo agoThis 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 3mo agoAnd 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.