BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//linuxsoftware.nz//NONSGML Joyous v1.4//EN
BEGIN:VEVENT
SUMMARY:Training optical neural networks for nonlinear logics: Towards ana
 log simulators of artificial life
DTSTART:20241121T150000Z
DTEND:20241121T170000Z
DTSTAMP:20260921T195435Z
UID:cbcf5829-4b04-4a82-bee1-ad5e3a0df435
SEQUENCE:1
CREATED:20241119T105259Z
DESCRIPTION:This research investigates the potential of optical systems to
  perform nonlinear computations using light\, an inherently linear medium.
  By introducing nonlinearity into a physical neural network\, we explore m
 odulation strategies to identify the optimal architecture for the problem 
 at hand. The initial phase involves modeling nonlinear material layers and
  analyzing their integration properties. As a benchmark\, we implement an 
 optical decoder and an optical AND gate using Laguerre-Gaussian vortex bea
 ms and Gaussian beams to encode binary states. The phase masks within the 
 system are designed with an adjoint algorithm\, termed wavefront matching\
 , which enables multiple input beams to interact and produce desired compu
 tational outputs based on predefined mapping rules. To further enhance the
  system&#x27\;s performance\, optimization is conducted in Fourier space. 
 To handle increasingly complex tasks\, we developed a parallel machine lea
 rning model to improve training efficiency and leverage gradient descent i
 n discovering the optimal architecture for the continuous properties of op
 tical components. Findings from this research confirm the feasibility of t
 raining optical neural networks for nonlinear computations\, as well as ad
 vancing coherent light manipulation.Beyond simple optical gates\, this app
 roach extends to simulating cellular automata (CA)\, typically created fro
 m simple rule sets. We investigate the system&#x27\;s ability to character
 ize state-update rules\, laying the groundwork for future applications\, s
 uch as translating Lenia — a continuous 2D cellular automaton model of a
 rtificial life — into an optical format. These advancements provide a fo
 undation not only for optical computing but also for a novel framework to 
 explore artificial life and cellular automata within resonating physical s
 ystems.
LAST-MODIFIED:20241119T105259Z
LOCATION:Advanced Training Room Physics Building - 2nd floor/Online
URL:http://df.vps.tecnico.ulisboa.pt/en/events/training-optical-neural-net
 works-for-nonlinear-logics-towards-analog-simulators-of-artificial-life/
X-ALT-DESC;FMTTYPE=text/html:<p data-block-key="47lrn">This research inves
 tigates the potential of optical systems to perform nonlinear computations
  using light\, an inherently linear medium. By introducing nonlinearity in
 to a physical neural network\, we explore modulation strategies to identif
 y the optimal architecture for the problem at hand. The initial phase invo
 lves modeling nonlinear material layers and analyzing their integration pr
 operties.<br/><br/> As a benchmark\, we implement an optical decoder and a
 n optical AND gate using Laguerre-Gaussian vortex beams and Gaussian beams
  to encode binary states. The phase masks within the system are designed w
 ith an adjoint algorithm\, termed wavefront matching\, which enables multi
 ple input beams to interact and produce desired computational outputs base
 d on predefined mapping rules.<br/><br/> To further enhance the system&#x2
 7\;s performance\, optimization is conducted in Fourier space. To handle i
 ncreasingly complex tasks\, we developed a parallel machine learning model
  to improve training efficiency and leverage gradient descent in discoveri
 ng the optimal architecture for the continuous properties of optical compo
 nents. Findings from this research confirm the feasibility of training opt
 ical neural networks for nonlinear computations\, as well as advancing coh
 erent light manipulation.<br/></p><p data-block-key="dcf64">Beyond simple 
 optical gates\, this approach extends to simulating cellular automata (CA)
 \, typically created from simple rule sets. We investigate the system&#x27
 \;s ability to characterize state-update rules\, laying the groundwork for
  future applications\, such as translating Lenia — a continuous 2D cellu
 lar automaton model of artificial life — into an optical format. These a
 dvancements provide a foundation not only for optical computing but also f
 or a novel framework to explore artificial life and cellular automata with
 in resonating physical systems.</p>
END:VEVENT
END:VCALENDAR