BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Multi-Agent Reinforcement Learning for Nonlocal Games via Shared V
 QCs
DTSTART;VALUE=DATE-TIME:20260910T094000Z
DTEND;VALUE=DATE-TIME:20260910T101000Z
DTSTAMP;VALUE=DATE-TIME:20260922T191356Z
UID:indico-contribution-2028@events.ncbj.gov.pl
DESCRIPTION:Speakers: Dawid Mazur (AGH University of Krakow)\nNonlocal gam
 es provide a framework for studying cooperative strategies under different
  resource assumptions. These range from independent agents with no shared 
 resource\, through local hidden variable (LHV)\, to entangled quantum stat
 es\, allowing one to quantify the advantage of quantum strategies over cla
 ssical ones. Discovering optimal quantum strategies\, however\, remains a 
 non-trivial task. We propose a multi-agent reinforcement learning (MARL) a
 pproach with shared variational quantum circuits (VQCs) as a general-purpo
 se tool for automated discovery of such strategies. Prior to the game\, th
 e agents may agree on a strategy\, but once the game begins\, each agent r
 eceives a question and must provide an answer without communicating. An en
 tangled state is prepared across the agents' respective registers. Each ag
 ent then applies local parameterized unitary operations conditioned on the
 ir received question\, followed by a measurement on their subsystem. The m
 easurement outcomes determine the agents' actions\, and the circuit is tra
 ined via the REINFORCE algorithm. We evaluate the method on three nonlocal
  games: the CHSH game\, the Magic Square game\, and the Rendezvous problem
 . Our results demonstrate that the proposed approach consistently exceeds 
 LHV bounds and achieves winning probabilities matching quantum optimal lim
 its.\n\nhttps://events.ncbj.gov.pl/event/468/contributions/2028/
LOCATION:
URL:https://events.ncbj.gov.pl/event/468/contributions/2028/
END:VEVENT
END:VCALENDAR