Saved in:
Bibliographic Details
Main Authors: Chen, Kuan-Cheng, Yu, Shang, Liu, Chen-Yu, Chen, Samuel Yen-Chi, Tseng, Huan-Hsin, Chang, Yen Jui, Huang, Wei-Hao, Burt, Felix, Gomez, Esperanza Cuenca, Chandani, Zohim, Clements, William, Walmsley, Ian, Leung, Kin K.
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.14898
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914396187394048
author Chen, Kuan-Cheng
Yu, Shang
Liu, Chen-Yu
Chen, Samuel Yen-Chi
Tseng, Huan-Hsin
Chang, Yen Jui
Huang, Wei-Hao
Burt, Felix
Gomez, Esperanza Cuenca
Chandani, Zohim
Clements, William
Walmsley, Ian
Leung, Kin K.
author_facet Chen, Kuan-Cheng
Yu, Shang
Liu, Chen-Yu
Chen, Samuel Yen-Chi
Tseng, Huan-Hsin
Chang, Yen Jui
Huang, Wei-Hao
Burt, Felix
Gomez, Esperanza Cuenca
Chandani, Zohim
Clements, William
Walmsley, Ian
Leung, Kin K.
contents Photonic quantum processors naturally produce intrinsically stochastic measurement outcomes, offering a hardware-native source of structured randomness that can be exploited during machine-learning training. Here we introduce Photonic Quantum-Enhanced Knowledge Distillation (PQKD), a hybrid quantum photonic--classical framework in which a programmable photonic circuit generates a compact conditioning signal that constrains and guides a parameter-efficient student network during distillation from a high-capacity teacher. PQKD replaces fully trainable convolutional kernels with dictionary convolutions: each layer learns only a small set of shared spatial basis filters, while sample-dependent channel-mixing weights are derived from shot-limited photonic features and mapped through a fixed linear transform. Training alternates between standard gradient-based optimisation of the student and sampling-robust, gradient-free updates of photonic parameters, avoiding differentiation through photonic hardware. Across MNIST, Fashion-MNIST and CIFAR-10, PQKD traces a controllable compression--accuracy frontier, remaining close to teacher performance on simpler benchmarks under aggressive convolutional compression. Performance degrades predictably with finite sampling, consistent with shot-noise scaling, and exponential moving-average feature smoothing suppresses high-frequency shot-noise fluctuations, extending the practical operating regime at moderate shot budgets.
format Preprint
id arxiv_https___arxiv_org_abs_2603_14898
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Photonic Quantum-Enhanced Knowledge Distillation
Chen, Kuan-Cheng
Yu, Shang
Liu, Chen-Yu
Chen, Samuel Yen-Chi
Tseng, Huan-Hsin
Chang, Yen Jui
Huang, Wei-Hao
Burt, Felix
Gomez, Esperanza Cuenca
Chandani, Zohim
Clements, William
Walmsley, Ian
Leung, Kin K.
Quantum Physics
Emerging Technologies
Machine Learning
Photonic quantum processors naturally produce intrinsically stochastic measurement outcomes, offering a hardware-native source of structured randomness that can be exploited during machine-learning training. Here we introduce Photonic Quantum-Enhanced Knowledge Distillation (PQKD), a hybrid quantum photonic--classical framework in which a programmable photonic circuit generates a compact conditioning signal that constrains and guides a parameter-efficient student network during distillation from a high-capacity teacher. PQKD replaces fully trainable convolutional kernels with dictionary convolutions: each layer learns only a small set of shared spatial basis filters, while sample-dependent channel-mixing weights are derived from shot-limited photonic features and mapped through a fixed linear transform. Training alternates between standard gradient-based optimisation of the student and sampling-robust, gradient-free updates of photonic parameters, avoiding differentiation through photonic hardware. Across MNIST, Fashion-MNIST and CIFAR-10, PQKD traces a controllable compression--accuracy frontier, remaining close to teacher performance on simpler benchmarks under aggressive convolutional compression. Performance degrades predictably with finite sampling, consistent with shot-noise scaling, and exponential moving-average feature smoothing suppresses high-frequency shot-noise fluctuations, extending the practical operating regime at moderate shot budgets.
title Photonic Quantum-Enhanced Knowledge Distillation
topic Quantum Physics
Emerging Technologies
Machine Learning
url https://arxiv.org/abs/2603.14898