6c21a3c1d966793ce3e5 · Multiplexed gradient descent: Fast online training of modern datasets on hardware neural networks without backpropagation
Multiplexed gradient descent: Fast online training of modern datasets on hardware neural networks without backpropagation
We show that model-free perturbative methods can be used to efficiently train modern neural network architectures in a way that can be directly applied to emerging neuromorphic hardware. These methods were investigated for training VLSI neural networks
{
"title_en": "Multiplexed gradient descent: Fast online training of modern datasets on hardware neural networks without backpropagation",
"designation": "",
"series": "",
"status": "",
"date": "2023-06-26",
"date_original": "June 26, 2023",
"url": "https://www.nist.gov/publications/multiplexed-gradient-descent-fast-online-training-modern-datasets-hardware-neural",
"source": "NIST Publications",
"display_en": "",
"authors_en": "Adam McCaughan , Bakhrom Oripov , Natesh Ganesh, Sae Woo Nam , Andrew Dienstfrey , Sonia Buckley",
"abstract_excerpt_en": "We show that model-free perturbative methods can be used to efficiently train modern neural network architectures in a way that can be directly applied to emerging neuromorphic hardware. These methods were investigated for training VLSI neural networks",
"id": "6c21a3c1d966793ce3e5",
"sources": [
"NIST Publications"
],
"source_records": [
{
"title_en": "Multiplexed gradient descent: Fast online training of modern datasets on hardware neural networks without backpropagation",
"designation": "",
"series": "",
"status": "",
"date": "2023-06-26",
"date_original": "June 26, 2023",
"url": "https://www.nist.gov/publications/multiplexed-gradient-descent-fast-online-training-modern-datasets-hardware-neural",
"source": "NIST Publications",
"display_en": "",
"authors_en": "Adam McCaughan , Bakhrom Oripov , Natesh Ganesh, Sae Woo Nam , Andrew Dienstfrey , Sonia Buckley",
"abstract_excerpt_en": "We show that model-free perturbative methods can be used to efficiently train modern neural network architectures in a way that can be directly applied to emerging neuromorphic hardware. These methods were investigated for training VLSI neural networks"
}
],
"translation_state": "none"
}