Beta. Content is under active construction and has not been peer-reviewed. Report errors on GitHub.Disclaimer

Representation Learning Theory

1 questionsDifficulty 7-7View topic

Advanced

0 / 1

1 advancedAdapts to your performance

1 / 1

advanced (7/10)compare

The Information Bottleneck (IB) objective

min_{encoder} I (X; Z) - β \cdot I (Z; Y)

was proposed as an explanation for deep learning generalization: networks allegedly compress the input during a distinct training phase. What is the main critique of this explanation?