@kaia I asked it to prove the spectral theorem at one point and it gave me an answer that started by assuming the hardest part of the theorem the ln followed it with a calculation that was incorrect but involved the same symbols and steps as the correct proof, just kind of jarbled.