4 ms·

I use LLMs more in the context of peer-reviewing and also came to a similar conclusion, gpt-5.5 codex xhigh reasoning seemed to catch more edge cases and went "

• Razengan 4mo ago
  Claude often spent most of its output listing all the things that were already correct and working! "This is good" and most of its findings were false positive or outright wrong as in the screenshot I posted above.