Alex Krizhevsky

scientist · 2 mentions across 2 readings

In this course

Krizhevsky is a co-author of AlexNet, the foundational deep convolutional neural network that catalyzed the modern deep learning revolution in computer vision (though his name appears only peripherally in these excerpts). The readings invoke his work implicitly when discussing batch normalization and sampling strategies in neural networks, suggesting his architectures established the baseline technical practices—like batch statistics at inference time—that contemporary generative models and classifiers now debate and refine. His influence underwrites the entire premise that neural networks' universality makes them capable of approximating any function, a theoretical guarantee that justifies their application across art, design, and cybernetic systems.

Mentioned in 2 readings

Andrew Brock

et al., LARGE SCALE GAN TRAINING FOR HIGH FIDELITY NATURAL IMAGE SYNTHESIS

The default behavior with batch normalized classifier networks is to use a running average of the activation moments at test time. Previous works (Radford et al., 2016) have instead used batch statistics when sampling images. While this is …

Dan McQuillan

from Resisting AI, an Anti-fascist Approach to Artificial Intelligence

The mathematical power of neural networks comes from their universality: in other words, for any input they can approximate the desired output function. As dry as ‘being able to compute any function’ may sound, it becomes a lot more compell…

Appears alongside

People mentioned in the same passages — sorted by co-occurrence weight.

Alec Radford 1 Geoffrey Hinton 1 Michael Nielsen 1