Inference and Gradient Measurement for Backpropagation in Photonic Neural Networks

Sunil Pai; Tyler W. Hughes; Taewon Park; Ben Bartlett; Ian Williamson; Momchil Minkov; Maziyar Milanizadeh; Nathnael Abebe; Francesco Morichetti; Andrea Melloni; Olav Solgaard; Shanhui Fan; David A.B. Miller

Inference and Gradient Measurement for Backpropagation in Photonic Neural Networks

Sunil Pai, Tyler W. Hughes, Taewon Park, Ben Bartlett, Ian Williamson, Momchil Minkov, Maziyar Milanizadeh, Nathnael Abebe, Francesco Morichetti, Andrea Melloni, Olav Solgaard, Shanhui Fan, David A.B. Miller

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

We experimentally demonstrate in situ backpropagation in a programmable nanophotonic interferometer network, achieving inference accuracies matching digital implementations. Error gradients are computed by simultaneously measuring optical interference at intermediate network components, eliminating expensive digital computations.

Original language	English
Title of host publication	2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781957171050
Publication status	Published - 2022
Externally published	Yes
Event	2022 Conference on Lasers and Electro-Optics, CLEO 2022 - San Jose, United States Duration: 15 May 2022 → 20 May 2022

Publication series

Name	2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings

Conference

Conference	2022 Conference on Lasers and Electro-Optics, CLEO 2022
Country/Territory	United States
City	San Jose
Period	15/05/22 → 20/05/22

Cite this

Pai, S., Hughes, T. W., Park, T., Bartlett, B., Williamson, I., Minkov, M., Milanizadeh, M., Abebe, N., Morichetti, F., Melloni, A., Solgaard, O., Fan, S., & Miller, D. A. B. (2022). Inference and Gradient Measurement for Backpropagation in Photonic Neural Networks. In 2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings Article STh5G.2 (2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings). Institute of Electrical and Electronics Engineers Inc..

@inproceedings{1698699b493e4677baf275676e22bce0,

title = "Inference and Gradient Measurement for Backpropagation in Photonic Neural Networks",

abstract = "We experimentally demonstrate in situ backpropagation in a programmable nanophotonic interferometer network, achieving inference accuracies matching digital implementations. Error gradients are computed by simultaneously measuring optical interference at intermediate network components, eliminating expensive digital computations.",

author = "Sunil Pai and Hughes, {Tyler W.} and Taewon Park and Ben Bartlett and Ian Williamson and Momchil Minkov and Maziyar Milanizadeh and Nathnael Abebe and Francesco Morichetti and Andrea Melloni and Olav Solgaard and Shanhui Fan and Miller, {David A.B.}",

note = "Publisher Copyright: {\textcopyright} Optica Publishing Group 2022, {\textcopyright} 2022 The Author(s); 2022 Conference on Lasers and Electro-Optics, CLEO 2022 ; Conference date: 15-05-2022 Through 20-05-2022",

year = "2022",

language = "English",

series = "2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings",

}

Pai, S, Hughes, TW, Park, T, Bartlett, B, Williamson, I, Minkov, M, Milanizadeh, M, Abebe, N, Morichetti, F, Melloni, A, Solgaard, O, Fan, S & Miller, DAB 2022, Inference and Gradient Measurement for Backpropagation in Photonic Neural Networks. in 2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings., STh5G.2, 2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings, Institute of Electrical and Electronics Engineers Inc., 2022 Conference on Lasers and Electro-Optics, CLEO 2022, San Jose, United States, 15/05/22.

Inference and Gradient Measurement for Backpropagation in Photonic Neural Networks. / Pai, Sunil; Hughes, Tyler W.; Park, Taewon et al.
2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2022. STh5G.2 (2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Inference and Gradient Measurement for Backpropagation in Photonic Neural Networks

AU - Pai, Sunil

AU - Hughes, Tyler W.

AU - Park, Taewon

AU - Bartlett, Ben

AU - Williamson, Ian

AU - Minkov, Momchil

AU - Milanizadeh, Maziyar

AU - Abebe, Nathnael

AU - Morichetti, Francesco

AU - Melloni, Andrea

AU - Solgaard, Olav

AU - Fan, Shanhui

AU - Miller, David A.B.

PY - 2022

Y1 - 2022

N2 - We experimentally demonstrate in situ backpropagation in a programmable nanophotonic interferometer network, achieving inference accuracies matching digital implementations. Error gradients are computed by simultaneously measuring optical interference at intermediate network components, eliminating expensive digital computations.

AB - We experimentally demonstrate in situ backpropagation in a programmable nanophotonic interferometer network, achieving inference accuracies matching digital implementations. Error gradients are computed by simultaneously measuring optical interference at intermediate network components, eliminating expensive digital computations.

UR - http://www.scopus.com/inward/record.url?scp=85139975739&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85139975739

T3 - 2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings

BT - 2022 Conference on Lasers and Electro-Optics, CLEO 2022 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2022 Conference on Lasers and Electro-Optics, CLEO 2022

Y2 - 15 May 2022 through 20 May 2022

ER -

Inference and Gradient Measurement for Backpropagation in Photonic Neural Networks

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this