William Newsome 教授を囲んでのワークショップと第4回注意研究会

****************************************
William Newsome 教授を囲んでのワークショップと
　　　　　　第4回注意研究会
****************************************
１．William Newsome スタンフォード大学教授を囲んでのワークショップ
開催日時：2009年11月25日（水）　13:00より
開催場所：京都大学　稲盛財団記念館　３階　大会議室
　　　 （荒神橋東詰にある、川端通に面した南北に細長い建物です）
プログラム：下記の問い合わせ先に問い合わせてください
２．第４回注意研究会
開催日時：平成21年11月25日（水曜日）　午後４時半より
開催場所：京都大学　稲盛財団記念館　３階　大会議室
　　　 （荒神橋東詰にある、川端通に面した南北に細長い建物です）
話題提供：William T. Newsome, Ph.D.
（Professor, Department of Neurobiology, Stanford University School of Medicine）
講演タイトル：
“Linking action to reward: multi-electrode array recordings from prefrontal cortex”
講演概要：
To investigate the neural basis of decision-making, we conducted multielectrode array recordings (96 electrodes) from the prearcuate gyrus while rhesus monkeys performed a motion discrimination task. We observed a wide range of visual, memory, and oculomotor signals as reported originally in the classic work of Funahashi and colleagues (J Neurophysiol, 1989), and we observed decision-related signals as reported originally by Kim and Shadlen (Nat Neurosci, 1999). A unique feature of our behavioral paradigm is that the reward was delivered after a variable 1-2 second delay after the monkey’s operant response. Surprisingly, the most prominent neural signals we observed in the prearcuate gyrus occurred during this reward-delay period, and appeared to represent a memory of the action executed by the monkey before receiving the reward. These action-memory signals may play an important role in learning by linking a current action to rewards that occur seconds (or perhaps longer) in!
to the future. Standard forms of spike-timing dependent plasticity cannot mediate learning under these circumstances because they only operate over tens of milliseconds at best. Action-memory signals in prefrontal cortex may provide the necessary link between current actions and future rewards, consistent with reinforcement learning theory.

2009/07/08