4 ms·
> The case in point is that most LLMs, including GPT-4o, cannot tell whether 9.11 or 9.8 is bigger! Wrong. GPT-4o gives me the correct answer to this question,
by meisel 2y ago
> The case in point is that most LLMs, including GPT-4o, cannot tell whether 9.11 or 9.8 is bigger!
Wrong. GPT-4o gives me the correct answer to this question, 9.8.
- orangecat 2y agoEven GPT-3.5 was correct, and so was Claude Sonnet 3.5. Haiku usually gets it wrong.