{"id":139275,"date":"2022-05-13T00:23:24","date_gmt":"2022-05-13T05:23:24","guid":{"rendered":"https:\/\/lifeboat.com\/blog\/2022\/05\/a-generalist-agent"},"modified":"2022-05-13T00:23:24","modified_gmt":"2022-05-13T05:23:24","slug":"a-generalist-agent","status":"publish","type":"post","link":"https:\/\/lifeboat.com\/blog\/2022\/05\/a-generalist-agent","title":{"rendered":"A Generalist Agent"},"content":{"rendered":"<p><a class=\"aligncenter blog-photo\" href=\"https:\/\/lifeboat.com\/blog.images\/a-generalist-agent3.jpg\"><\/a><\/p>\n<p>Inspired by progress in large-scale language modelling, we apply a similar approach towards building a single generalist agent beyond the realm of text outputs. The agent, which we refer to as Gato, works as a multi-modal, multi-task, multi-embodiment generalist policy. The same network with the same weights can play Atari, caption images, chat, stack blocks with a real robot arm and much more, deciding based on its context whether to output text, joint torques, button presses, or other tokens.<\/p>\n<p>During the training phase of Gato, data from different tasks and modalities are serialised into a flat sequence of tokens, batched, and processed by a transformer neural network similar to a large language model. The loss is masked so that Gato only predicts action and text targets.<\/p>\n<p>When deploying Gato, a prompt, such as a demonstration, is tokenised, forming the initial sequence. Next, the environment yields the first observation, which is also tokenised and appended to the sequence. Gato samples the action vector autoregressively, one token at a time.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Inspired by progress in large-scale language modelling, we apply a similar approach towards building a single generalist agent beyond the realm of text outputs. The agent, which we refer to as Gato, works as a multi-modal, multi-task, multi-embodiment generalist policy. The same network with the same weights can play Atari, caption images, chat, stack blocks [\u2026]<\/p>\n","protected":false},"author":359,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[31,6],"tags":[],"class_list":["post-139275","post","type-post","status-publish","format-standard","hentry","category-policy","category-robotics-ai"],"_links":{"self":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/139275","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/users\/359"}],"replies":[{"embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/comments?post=139275"}],"version-history":[{"count":0,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/139275\/revisions"}],"wp:attachment":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/media?parent=139275"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/categories?post=139275"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/tags?post=139275"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}