i think of myself as a LLM and recently i think my next token prediction is getting really good, either because of some scaling law, or that everyone around me is trained with the same weights
136