BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:UW-Madison-Physics-Events
BEGIN:VEVENT
SEQUENCE:1
UID:UW-Physics-Event-5428
DTSTART:20200617T160000Z
DTEND:20200617T170000Z
DTSTAMP:20260415T023619Z
LAST-MODIFIED:20200428T224048Z
LOCATION:Please register for this online event: http://physicsmeetsml.
 org
SUMMARY:Deep Learning and Quantum Gravity\, Physics ∩ ML Seminar\, K
 oji Hashimoto\, Osaka University
DESCRIPTION:Formulating quantum gravity is one of the final goals of f
 undamental physics. Recent progress in string theory brought a concret
 e formulation called AdS/CFT correspondence\, in which a gravitational
  spacetime emerges from lower-dimensional non gravitational quantum sy
 stems\, but we still lack in understanding how the correspondence work
 s. I discuss similarities between the quantum gravity and deep learnin
 g architecture\, by regarding the neural network as a discretized spac
 etime. In particular\, the questions such as\, when\, why and how a ne
 ural network can be a space or a spacetime\, may lead to a novel way t
 o look at machine learning. I implement concretely the AdS/CFT framewo
 rk into a deep learning architecture\, and show the emergence of a cur
 ved spacetime as a neural network\, from a given teacher data of quant
 um systems.
URL:https://www.physics.wisc.edu/events/?id=5428
END:VEVENT
END:VCALENDAR