重大X投稿2026-10-02 15:57 Introducing Gacha Decoding 🎲: we show how to diversify LM responses, beating prior work by >11x in sample efficiency. Reroll your LM for va… 報道ソース Nathan Lambert2026-10-02 15:57