BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Computational Neuroscience Journal Club - Xizi Li (CBL)
DTSTART:20180612T150000Z
DTEND:20180612T160000Z
UID:TALK106864@talks.cam.ac.uk
CONTACT:Rodrigo Echeveste
DESCRIPTION:Xizi Li will cover:\n\n• Vector-based navigation using grid-
 like representations in artificial agents\n\n• Banino et al (DeepMind)\n
 \n• Nature (2018)\n\n• https://www.nature.com/articles/s41586-018-0102
 -6\n\nAbstract: Deep neural networks have achieved impressive successes in
  fields ranging from object recognition to complex games such as Go. Navig
 ation\, however\, remains a substantial challenge for artificial agents\, 
 with deep neural networks trained by reinforcement learning failing to riv
 al the proficiency of mammalian spatial behaviour\, which is underpinned b
 y grid cells in the entorhinal cortex. Grid cells are thought to provide a
  multi-scale periodic representation that functions as a metric for coding
  space and is critical for integrating self-motion (path integration) and 
 planning direct trajectories to goals (vector-based navigation). Here we s
 et out to leverage the computational functions of grid cells to develop a 
 deep reinforcement learning agent with mammal-like navigational abilities.
  We first trained a recurrent network to perform path integration\, leadin
 g to the emergence of representations resembling grid cells\, as well as o
 ther entorhinal cell types. We then showed that this representation provid
 ed an effective basis for an agent to locate goals in challenging\, unfami
 liar\, and changeable environments—optimizing the primary objective of n
 avigation through deep reinforcement learning. The performance of agents e
 ndowed with grid-like representations surpassed that of an expert human an
 d comparison agents\, with the metric quantities necessary for vector-base
 d navigation derived from grid-like units within the network. Furthermore\
 , grid-like representations enabled agents to conduct shortcut behaviours 
 reminiscent of those performed by mammals. Our findings show that emergent
  grid-like representations furnish agents with a Euclidean spatial metric 
 and associated vector operations\, providing a foundation for proficient n
 avigation. As such\, our results support neuroscientific theories that see
  grid cells as critical for vector-based navigation\, demonstrating that t
 he latter can be combined with path-based strategies to support navigation
  in challenging environments.\n
LOCATION:Cambridge University Engineering Department\, CBL\, BE4-38 (http:
 //learning.eng.cam.ac.uk/Public/Directions)
END:VEVENT
END:VCALENDAR