{"id":1490,"date":"2019-11-06T20:54:00","date_gmt":"2019-11-06T11:54:00","guid":{"rendered":"http:\/\/cinetjp-static3.nict.go.jp\/japanese\/?p=1490"},"modified":"2022-08-27T09:48:03","modified_gmt":"2022-08-27T00:48:03","slug":"20191015_4201","status":"publish","type":"event","link":"http:\/\/cinetjp-static3.nict.go.jp\/japanese\/event\/20191015_4201\/","title":{"rendered":"40th CiNet Monthly Seminar: Andrew Jackson \u201cWhile you were sleeping \u2013 A hypothesis about off-line learning\u201d"},"content":{"rendered":"\n
CiNet Monthly Seminar<\/p>\n\n\n\n
2019\u5e7411\u67086\u65e5\uff08\u6c34\uff09
16:00-17:00
\u4f1a\u5834\uff1a CiNet\u30001F\u3000\u5927\u4f1a\u8b70\u5ba4<\/p>\n\n\n\n
\u201cWhile you were sleeping \u2013 A hypothesis about off-line learning and the cerebellum\u201d<\/p>\n\n\n\n
Andrew Jackson
Newcastle University<\/p>\n\n\n\n
Host :\u00a0\u7fbd\u5009 \u4fe1\u5b8f<\/a>\uff08\u6c60\u8c37 G<\/a>\uff09<\/p>\n\n\n\n Abstract: <\/p>\n","protected":false},"featured_media":0,"template":"","acf":[],"_links":{"self":[{"href":"http:\/\/cinetjp-static3.nict.go.jp\/japanese\/wp-json\/wp\/v2\/event\/1490"}],"collection":[{"href":"http:\/\/cinetjp-static3.nict.go.jp\/japanese\/wp-json\/wp\/v2\/event"}],"about":[{"href":"http:\/\/cinetjp-static3.nict.go.jp\/japanese\/wp-json\/wp\/v2\/types\/event"}],"wp:attachment":[{"href":"http:\/\/cinetjp-static3.nict.go.jp\/japanese\/wp-json\/wp\/v2\/media?parent=1490"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
Anecdotally, we know that a difficult problem can often be solved by \u2018sleeping on it\u2019, but the mechanisms of off-line learning are unclear. In this talk, I will discuss evidence from human behavioural experiments and wireless electrophysiological recording in non-human primates that suggests sleep-dependent performance gains derive from cerebellar forward models. Put simply, during the day we learn \u2018if I do x, then y happens\u2019, while in sleep we invert this to learn \u2018if I want y to happen, then I should do x\u2019. This mechanism may be supported by coupled spindle oscillations mediating information transfer via cerebello-thalamo-cortical pathways, and represents a powerful, computationally-efficient solution to problems inherent in model-free reinforcement learning. Finally, I will speculate that this hypothesis may link observed learning deficits and reduced spindle density in conditions linked to cerebellar dysfunction including schizophrenia and autism.<\/p>\n\n\n\n