Alexandria Library

A curated collection of awesome articles

Here are some ressources related to Neural networks, Training, Learning. Hoping you will find learning material, alternative tools or just some interesting stuff.

Home > Programming, Machine Learning, Quan... > Machine Learning, Deep Learning, Ne... > Neuroscience, Neural Networks, Deep... > Neural networks, Training, Learning...

Articles

Ending an Ugly Chapter in Chip Design
See discussion (46 comments, 97 points) - Archive
Fourier analysis may help to quickly train more accurate neural networks
See discussion (35 comments, 105 points) - Archive
Researchers Discover a More Flexible Approach to Machine Learning
See discussion (32 comments, 155 points) - Archive
New neural network architecture inspired by neural system of a worm
See discussion (65 comments, 218 points) - Archive
Neural Networks and the Chomsky Hierarchy
See discussion (79 comments, 164 points) - Archive
Geoffrey Hinton publishes new deep learning algorithm
See discussion (121 comments, 319 points) - Archive
A Generalist Neural Algorithmic Learner
See discussion (15 comments, 92 points) - Archive
Meta Learning Backpropagation and Improving It (2020)
See discussion (29 comments, 138 points) - Archive
Neural Network Loss Landscapes: What do we know? (2021)
See discussion (23 comments, 152 points) - Archive
Techniques for Training Large Neural Networks
See discussion (23 comments, 133 points) - Archive
A modern self-referential weight matrix that learns to modify itself
See discussion (46 comments, 166 points) - Archive
Simple electrical circuit learns on its own–with no help from a computer
See discussion (45 comments, 130 points) - Archive
Neurosymbolic Programming (2021) [pdf]
See discussion (28 comments, 85 points) - Archive
Computer scientists prove why bigger neural networks do better
See discussion (145 comments, 282 points) - Archive
Deep physical neural networks trained with backpropagation
See discussion (36 comments, 108 points) - Archive
Researchers Build AI That Builds AI
See discussion (54 comments, 127 points) - Archive
Permutation-Invariant Neural Networks for Reinforcement Learning
See discussion (16 comments, 109 points) - Archive
Predictive coding has been unified with backpropagation
See discussion (85 comments, 311 points) - Archive
A New Lens on Understanding Generalization in Deep Learning
See discussion (35 comments, 91 points) - Archive
Understanding deep learning requires rethinking generalization
See discussion (40 comments, 165 points) - Archive
Don't Mess with Backprop: Doubts about Biologically Plausible Deep Learning
See discussion (53 comments, 91 points) - Archive
The neural network of the Stockfish chess engine
See discussion (94 comments, 355 points) - Archive
Researchers propose faster, more efficient alternatives to backpropagation
See discussion (103 comments, 239 points) - Archive
You might not need machine learning
See discussion (193 comments, 437 points) - Archive
What is protein folding? A brief explanation
See discussion (78 comments, 314 points) - Archive
Neural Databases
See discussion (30 comments, 124 points) - Archive
Predictive Coding Approximates Backprop Along Arbitrary Computation Graphs
See discussion (12 comments, 92 points) - Archive
Neural programmer better than Quicksort
See discussion (126 comments, 282 points) - Archive
Deep Molecular Programming
See discussion (11 comments, 130 points) - Archive
When to assume neural networks can solve a problem
See discussion (42 comments, 165 points) - Archive
Lagrangian Neural Networks
See discussion (49 comments, 246 points) - Archive
Proving the Lottery Ticket Hypothesis: Pruning is All You Need
See discussion (61 comments, 174 points) - Archive
Generative Teaching Networks: Accelerating Neural Architecture Search
See discussion (15 comments, 116 points) - Archive
Using neural networks to solve advanced mathematics equations
See discussion (21 comments, 101 points) - Archive
A watershed moment for protein structure prediction
See discussion (80 comments, 272 points) - Archive
Deep Learning Breakthrough Made by Rice University Scientists
See discussion (33 comments, 112 points) - Archive
Understanding the generalization of ‘lottery tickets’ in neural networks
See discussion (16 comments, 98 points) - Archive
Sample Efficient Evolutionary Algorithm for Analog Circuit Design
See discussion (25 comments, 114 points) - Archive
Exploring Weight Agnostic Neural Networks
See discussion (64 comments, 179 points) - Archive
Neural Networks Are Essentially Polynomial Regression
See discussion (40 comments, 122 points) - Archive
Exploring DNA with Deep Learning
See discussion (13 comments, 116 points) - Archive
Theories of Error Back-Propagation in the Brain
See discussion (33 comments, 160 points) - Archive
Weight-Agnostic Neural Networks
See discussion (37 comments, 158 points) - Archive
Neural nets typically contain smaller “subnetworks” that can often learn faster
See discussion (20 comments, 128 points) - Archive
An End-to-End AutoML Solution for Tabular Data at KaggleDays
See discussion (37 comments, 96 points) - Archive
Smarter Training of Neural Networks
See discussion (40 comments, 204 points) - Archive
Survey of Dropout Methods for Deep Neural Networks
See discussion (5 comments, 88 points) - Archive
Kicking neural network design automation into high gear
See discussion (3 comments, 92 points) - Archive
GradientFlow: Training ImageNet in 1.5 Minutes on 512 GPUs
See discussion (44 comments, 91 points) - Archive
Neural Networks Are Essentially Polynomial Regression (2018)
See discussion (75 comments, 173 points) - Archive
A Bayesian Perspective on Generalization and Stochastic Gradient Descent
See discussion (36 comments, 182 points) - Archive
Introducing AdaNet: Fast and Flexible AutoML with Learning Guarantees
See discussion (16 comments, 103 points) - Archive
Google's AutoML: Cutting Through the Hype
See discussion (52 comments, 155 points) - Archive
Program Synthesis in 2017-18
See discussion (10 comments, 98 points) - Archive
The Lottery Ticket Hypothesis: Finding Small, Trainable Neural Networks
See discussion (14 comments, 137 points) - Archive
AdamW and Super-convergence is now the fastest way to train neural nets
See discussion (37 comments, 329 points) - Archive
Polynomial Regression as an Alternative to Neural Nets
See discussion (46 comments, 137 points) - Archive
Neural Network Quine
See discussion (25 comments, 138 points) - Archive
Population-based training of neural networks
See discussion (22 comments, 188 points) - Archive
The quest to evolve neural networks through evolutionary algorithms
See discussion (104 comments, 213 points) - Archive
Neural Nets Can Learn Function Type Signatures from Binaries [pdf]
See discussion (43 comments, 213 points) - Archive
Scalable and Sustainable Deep Learning via Randomized Hashing
See discussion (4 comments, 92 points) - Archive
Understanding deep learning requires re-thinking generalization
See discussion (35 comments, 169 points) - Archive
Neural Architecture Search with Reinforcement Learning
See discussion (7 comments, 88 points) - Archive
Show HN: Deep Learning for Program Synthesis
See discussion (23 comments, 93 points) - Archive
Building Safe AI: A Tutorial on Homomorphically Encrypted Deep Learning
See discussion (95 comments, 228 points) - Archive
Enabling Continual Learning in Neural Networks
See discussion (19 comments, 152 points) - Archive
DeepCoder: Learning to Write Programs
See discussion (54 comments, 183 points) - Archive
Outrageously Large Neural Networks: The Sparsely-Gated Mixture-Of-Experts Layer
See discussion (81 comments, 200 points) - Archive
Neural Architecture Search with Reinforcement Learning
See discussion (38 comments, 116 points) - Archive
Exact mapping between Variational Renormalization Group and Deep Learning (2014)
See discussion (13 comments, 97 points) - Archive
Neural Networks for the Prediction of Organic Chemistry Reactions
See discussion (31 comments, 108 points) - Archive
Neural-redis – Neural networks module for Redis
See discussion (19 comments, 229 points) - Archive
One-shot Learning with Memory-Augmented Neural Networks
See discussion (19 comments, 103 points) - Archive
The Unreasonable Reputation of Neural Networks
See discussion (143 comments, 269 points) - Archive
Why Deep Learning Works II: the Renormalization Group
See discussion (33 comments, 120 points) - Archive
10B Parameter Neural Networks in Your Basement [pdf]
See discussion (15 comments, 114 points) - Archive
The Flaw Lurking in Every Deep Neural Net
See discussion (59 comments, 127 points) - Archive
DreamCoder: Growing Generalizable, Interpretable Knowledge
See discussion (7 comments, 83 points) - Archive
Show HN: Richard – A CNN written in C++ and Vulkan (no ML or math libs)
See discussion (24 comments, 184 points) - Archive
How do neural networks learn?
See discussion (129 comments, 214 points) - Archive
The History of Machine Learning in Trackmania
See discussion (23 comments, 123 points) - Archive
An Analog Network of Resistors Promises Machine Learning Without a Processor
See discussion (38 comments, 97 points) - Archive
Training of Physical Neural Networks
See discussion (46 comments, 142 points) - Archive
A new type of neural network is more interpretable
See discussion (86 comments, 337 points) - Archive