3 ms·
The author assumes that asking GPT to “take longer and reconsider” is something that an LLM actually does. As in, temporally. The author claims that “even aft
by maxdoop 4y ago
The author assumes that asking GPT to “take longer and reconsider” is something that an LLM actually does.
As in, temporally.
The author claims that “even after taking longer to reconsider” an answer, ChatGPT was still wrong.
The author appears to misunderstand how LLMs work.