Seriously, who at OpenAI is so stupid. First, I was getting every fith post warned when talking about IPv6 implementation details.
Then it went better. Great.
Now, 2 days later, I talk about differences between GPT and RTN architectures and I an not even submit because asking about a smaller chunk size and the math - VIOLATES THE CONTENT POLICY?
Have they not realizes yet that their review is more stupid than Llama2 ever was? IT. DOES. NOT. WORk. It creates more false positives than I have ever seen.