RT @ofirnachum: Say you have a bunch of logged data of agents (eg PPO) learning various RL tasks. How should you distill this data into a single agent that can quickly learn new tasks? Simple autoregressive modeling would give you a learner no better than the agents it trained from.... pic.twitter.com/1Wu44jbjJv
posted at 06:02:55
RT @Mikoryth: 文化庁のセミナー「AIと著作権」の内容を漫画にまとめてみました！ (1/3) #AIと著作権 #AIイラスト #漫画が読めるハッシュタグ pic.twitter.com/0QcHL3vDip
posted at 00:18:57