3 ms·
To clarify, by "doing the opposite" I mean OpenAI releasing GPT-4.5, a non-reasoning model that does worse on benchmarks (but supposed to be qualitatively bette
by andai 2y ago
To clarify, by "doing the opposite" I mean OpenAI releasing GPT-4.5, a non-reasoning model that does worse on benchmarks (but supposed to be qualitatively better). People shit on OpenAI hard for doing that.