BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Neural mechanisms of model-based planning in the rat - Kevin Mille
 r\, University College London
DTSTART:20181120T110000Z
DTEND:20181120T120000Z
UID:TALK115159@talks.cam.ac.uk
CONTACT:Marcelo Gomes Mattar
DESCRIPTION:Planning can be defined as use of an internal model\, containi
 ng knowledge about the outcomes likely to follow each possible action\, to
  guide action selection. In recent work\, we adapted for rodents a multi-s
 tep decision task widely used to study planning in human subjects\, allowi
 ng the extensive experimental toolkit available for rodents to be brought 
 to bear on this problem in a new way. We found that rats adopt a strategy 
 of model-based planning to solve this task\, and that silencing neural act
 ivity in either the orbitofrontal cortex or the dorsal hippocampus was suf
 ficient to impair planning. Here\, I will describe new data from experimen
 ts designed to reveal the computational role in model-based cognition play
 ed by each region. In the orbitofrontal cortex (OFC)\, neurons encode info
 rmation about expected outcomes in a manner specifically suitable for a ro
 le in model-based learning\, but not for a role in model-based choice. Tri
 al-by-trial optogenetic inactivations of the OFC similarly reveal a patter
 n of impairment that is consistent with impaired learning\, but not with i
 mpaired decision-making. These data suggest that the OFC supports model-ba
 sed cognition by signaling expected outcomes to a process which updates ch
 oice mechanisms residing elsewhere in the brain. In the dorsal hippocampus
 \, activity of CA1 neurons does not seem to encode information about expec
 ted outcomes\, but instead indexes the various behavioral states of the ta
 sk in a manner reminiscent of “place cell” coding. Sharp wave ripple e
 vents which occur during the inter-trial interval show evidence of “repl
 ay”\, in which sequences of neural activity typical of task performance 
 (occupying several seconds) are rapidly repeated within the span of a sing
 le sharp-wave ripple (occupying severals tens of milliseconds). Ongoing wo
 rk seeks to test whether the content of these replay events is consistent 
 with computational theories proposing roles in learning\, decision-making\
 , or both.
LOCATION:Cambridge University Engineering Department\, CBL\, BE4-38 (http:
 //learning.eng.cam.ac.uk/Public/Directions)
END:VEVENT
END:VCALENDAR