AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Struggling to Understand Equivariant and Invariant Concepts in Neural Networks
The narrator finds the work fascinating but struggles to understand it fully. They liken the process to trying to figure out a shape with eyes closed. After reading other papers, they find connections to the work and start to comprehend it more. They explain the difference between equivariant and invariant: in equivariant, shuffling the inputs also shuffles the output, while in invariant, shuffling the inputs doesn't change the output. They mention the use of convolutional neural networks for scanning images.