@_rockt : Great to see the next keynote by @proceduralia at the AutoRL workshop also using @NetHack_LE. Motif (arxiv.org/abs/23150.00166) uses LLMs to intrinsically motivate agents, leading to better score in NetHack than training on the environment's extrinsic reward 🤯 • TwiCopy

Tim Rocktäschel

@_rockt

+ Follow

Director and Open-Endedness Team Lead @GoogleDeepMind, Professor of AI @AI_UCL, PI @UCL_DARK, Fellow @ELLISforEurope.

ID: 228792418

linkhttp://rockt.ai calendar_today20-12-2010 17:56:54

5,5K Tweet

34,34K Followers

1,1K Following

Tim Rocktäschel

@_rockt

a year ago

Great to see the next keynote by Pierluca D'Oro at the AutoRL workshop also using The NetHack Learning Environment. Motif (arxiv.org/abs/2310.00166) uses LLMs to intrinsically motivate agents, leading to better score in NetHack than training on the environment's extrinsic reward 🤯

Great to see the next keynote by <a href="/proceduralia/">Pierluca D'Oro</a> at the AutoRL workshop also using <a href="/NetHack_LE/">The NetHack Learning Environment</a>. Motif (arxiv.org/abs/2310.00166) uses LLMs to intrinsically motivate agents, leading to better score in NetHack than training on the environment's extrinsic reward 🤯

thumb_up_off_alt17

chat_bubble_outline1

repeat5

shareShare