Goto

Collaborating Authors

 in-distribution interpretability

In-Distribution Interpretability for Challenging Modalities

Heiß, Cosmas, Levie, Ron, Resnick, Cinjon, Kutyniok, Gitta, Bruna, Joan

arXiv.org Machine Learning

It is widely recognized that the predictions of deep neural networks are difficult to parse relative to simpler approaches. However, the development of methods to investigate the mode of operation of such models has advanced rapidly in the past few years. Recent work introduced an intuitive framework which utilizes generative models to improve on the meaningfulness of such explanations. In this work, we display the flexibility of this method to interpret diverse and challenging modalities: music and physical simulations of urban environments.

  artificial intelligence, in-distribution interpretability, machine learning, (17 more...)
2007.00758
  Country:
  Genre: Research Report (0.41)
  Technology: Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (0.34)

Add feedback