Nonlinear optical encoding enabled by recurrent linear scattering
Optical information processing and computing can potentially offer enhanced performance,
scalability and energy efficiency. However, achieving nonlinearity—a critical component of …
scalability and energy efficiency. However, achieving nonlinearity—a critical component of …
Physical deep learning with biologically inspired training method: gradient-free approach for physical hardware
Ever-growing demand for artificial intelligence has motivated research on unconventional
computation based on physical devices. While such computation devices mimic brain …
computation based on physical devices. While such computation devices mimic brain …
Linear optical random projections without holography
We introduce what we believe to be a novel method to perform linear optical random
projections without the need for holography. Our method consists of a computationally trivial …
projections without the need for holography. Our method consists of a computationally trivial …
Walking noise: Understanding implications of noisy computations on classification tasks
Machine learning methods like neural networks are extremely successful and popular in a
variety of applications, however, they come at substantial computational costs, accompanied …
variety of applications, however, they come at substantial computational costs, accompanied …
Walking Noise: On Layer-Specific Robustness of Neural Architectures Against Noisy Computations and Associated Characteristic Learning Dynamics
Deep neural networks are extremely successful in various applications, however they exhibit
high computational demands and energy consumption. This is exacerbated by stuttering …
high computational demands and energy consumption. This is exacerbated by stuttering …
Benchmarking the accuracy and robustness of feedback alignment algorithms
AJ Sanfiz, M Akrout - arXiv preprint arXiv:2108.13446, 2021 - arxiv.org
Backpropagation is the default algorithm for training deep neural networks due to its
simplicity, efficiency and high convergence rate. However, its requirements make it …
simplicity, efficiency and high convergence rate. However, its requirements make it …
Optical training of large-scale Transformers and deep neural networks with direct feedback alignment
Modern machine learning relies nearly exclusively on dedicated electronic hardware
accelerators. Photonic approaches, with low consumption and high operation speed, are …
accelerators. Photonic approaches, with low consumption and high operation speed, are …
Ropust: improving robustness through fine-tuning with photonic processors and synthetic gradients
Robustness to adversarial attacks is typically obtained through expensive adversarial
training with Projected Gradient Descent. Here we introduce ROPUST, a remarkably simple …
training with Projected Gradient Descent. Here we introduce ROPUST, a remarkably simple …
Scaling Laws Beyond Backpropagation
Alternatives to backpropagation have long been studied to better understand how biological
brains may learn. Recently, they have also garnered interest as a way to train neural …
brains may learn. Recently, they have also garnered interest as a way to train neural …
On Noise Stability and Robustness of Adversarially Trained Networks on NVM Crossbars
Applications based on deep neural networks (DNNs) have grown exponentially in the past
decade. To match their increasing computational needs, several nonvolatile memory (NVM) …
decade. To match their increasing computational needs, several nonvolatile memory (NVM) …