BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Deep Reinforcement Learning from Human Preferences - Jessica Yung 
 (University of Cambridge)
DTSTART:20171115T170000Z
DTEND:20171115T183000Z
UID:TALK95629@talks.cam.ac.uk
CONTACT:Adrià Garriga Alonso
DESCRIPTION:How do you teach an algorithm to do a backflip or play a game 
 where rewards are sparse? In this seminar we will discuss how algorithms c
 an learn from human preferences as opposed to from pre-specified goal func
 tions. \n\nRemoving the need for humans to write goal functions is importa
 nt because getting them slightly wrong could lead to dangerous behaviour. 
 Here this is only used to learn physical behaviours\, but one can imagine 
 that it could apply to learning moral values as well. \n\nWe will be looki
 ng at the paper ‘Deep Reinforcement Learning from Human Preferences’ (
 Christiano et. al.\, 2017). We will discuss the model used and experiments
  in three domains: simulated robotics\, Atari arcade games and novel behav
 iours.\n\nLink to paper: https://arxiv.org/abs/1706.03741\n\nSlides: https
 ://valuealignment.ml/talks/2017-11-15-deeprl-human-prefs.pdf
LOCATION: Cambridge University Engineering Department\, CBL Seminar room B
 E4-38.  For directions see http://learning.eng.cam.ac.uk/Public/Directions
END:VEVENT
END:VCALENDAR