anyone want to guess parameter sizes here for GPT‑4.1, GPT‑4.1 mini GPT‑4.1 nano I'll start with 800 bn MoE (probably 120 bn activated), 200 bn MoE (33 bn

anyone want to guess parameter sizes here for GPT‑4.1, GPT‑4.1 mini GPT‑4.1 nano I'll start with 800 bn MoE (probably 120 bn activated), 200 bn MoE (33 bn activated), and 7bn parameter for nano