MNIST database
PulseAugur coverage of MNIST database — every cluster mentioning MNIST database across labs, papers, and developer communities, ranked by signal.
- used by alphaXiv 90%
- instance of The Street View House Numbers Dataset 90%
- used by Quantum Machine Learning 90%
- used by Spiking neural networks 80%
- used by Variational Autoencoders 80%
- used by dSprites 80%
- used by CIFAR-10 70%
- instance of CIFAR-10 70%
- used by Fashion-MNIST 70%
- instance of Fashion-MNIST 70%
- used by ScienceCast 70%
- used by Gotit.pub 70%
20 day(s) with sentiment data
-
New continuous-state probabilistic computer hardware demonstrated
Researchers have demonstrated a novel continuous-state probabilistic computer (CPC) that utilizes programmable energy-defined sampling for complex problem-solving. Unlike traditional binary systems, this CPC handles con…
-
New generative model framework uses nonlinear Schrödinger systems
Researchers have introduced a novel generative modeling framework that leverages forward-backward partial differential equations (PDEs), drawing connections to mean-field game theory and nonlinear Schrödinger systems. T…
-
Brain-Inspired Engram Memory Network Enhances AI Explainability
Researchers have developed an Engram Memory Network (EMN), a novel AI model inspired by the human memory system, to improve the explainability and reliability of prototype-based AI. The EMN utilizes a Hopfield-inspired …
-
New Target-Response Jacobian metric outperforms influence measures in AI model analysis
Researchers have introduced the target-response Jacobian ($J$), a new metric for understanding how model predictions change in response to alterations in individual training targets. This metric complements existing tra…
-
New framework reveals blind spots in AI distribution shift detection
Researchers have introduced a new framework called Conformal Test Martingales (CTMs) designed to detect distribution shifts in data. These CTMs utilize p-values generated by a conformity function, but they can be blind …
-
New research explores ReLU networks beyond linear separability
Researchers have analyzed a simplified neural network model with ReLU activation functions and logistic loss on non-linearly separable data. They identified an 'active cone' in parameter space where perfect classifiers …
-
StratFL combats covert model poisoning in federated learning
Researchers have developed StratFL, a novel federated learning strategy designed to counter covert model poisoning attacks. These attacks, often employed by advanced persistent threats, involve intermittently injecting …
-
FuzzInsight improves DNN fuzzing with new sampling and coverage strategies
Researchers have developed FuzzInsight, a new framework designed to improve the efficiency and effectiveness of fuzzing techniques for Deep Neural Networks (DNNs). The system addresses bottlenecks in seed utilization an…
-
New CIIR framework connects diverse datasets without paired supervision
Researchers have developed a novel framework called CIIR (Class-Indexed Interface Representation Learning) designed to connect heterogeneous datasets without requiring synchronized or instance-matched samples. This modu…
-
New SDFL method boosts robustness against label-flipping attacks
Researchers have developed a new method for semi-decentralized federated learning (SDFL) designed to enhance robustness against label-flipping attacks. This approach, detailed in a recent arXiv paper, combines a robust …
-
New criterion optimizes weight initialization for deep neural networks
Researchers have developed a new mathematical criterion for selecting the optimal variance of weight initialization in deep neural networks (DNNs). By analyzing stochastic gradient descent (SGD) through a continuous-tim…
-
New TASTE framework enhances out-of-distribution detection using Stein operators
Researchers have introduced TASTE (Task-Aware STEin operators), a novel framework for out-of-distribution (OOD) detection in machine learning. Unlike existing methods that focus solely on input data or classifier output…
-
New PAC-Bayes Theory Achieves Dimension-Free Generalization for Deep Neural Networks
Researchers have developed a new PAC-Bayesian generalization theory for deep neural networks that is independent of network dimension. This approach shifts prior specification from internal parameters to the non-paramet…
-
Weight perturbation approaches backpropagation accuracy on MNIST
A new research paper explores the potential of weight perturbation (WP) as an alternative to backpropagation (BP) for training neural networks. The study demonstrates that WP can achieve accuracy levels close to BP on t…
-
New research reveals 'accuracy well' phenomenon in deep learning dynamics
Researchers have identified a phenomenon in deep learning models called the "accuracy well," where accuracy initially increases, then decreases, before finally recovering. This non-monotone behavior has been observed ac…
-
Kolmogorov-Arnold Networks: Benefits questioned, educational tool released
A new paper questions the universal benefits of Kolmogorov-Arnold Networks (KANs), suggesting their favorable scaling properties may be tied to specific target structures rather than an inherent architectural advantage.…
-
New polynomial activation enables deep networks to train without normalization
Researchers have developed a novel polynomial activation function, f(z) = z + λz^3, which can be clamped and projected onto the unit sphere. This activation function allows deep neural networks to train effectively up t…
-
New Hierarchical Bayesian Method Enhances AI Uncertainty Quantification
Researchers have developed a new method called multilevel particle marginal Metropolis-Hastings (MLPMMH) to improve uncertainty quantification in AI by using Hierarchical Bayesian deep learning. This method combines pse…
-
Delayed generalization observed in linear models, study finds
Researchers have demonstrated that delayed generalization, a phenomenon where training accuracy significantly outpaces test accuracy, can occur even in simplified linear models. This effect arises when using gradient de…
-
Forward-Only Deep Learning Algorithms Show Mixed Energy Efficiency
A new study published on arXiv benchmarks forward-only deep learning algorithms against traditional backpropagation. The research found that while some forward-only methods like Forward-Forward and Cascaded Forward are …