Yoshua Bengio

TL;DR: A regularization method is proposed, GradMask, which penalizes saliency maps inferred from the classifier gradients when they are not consistent with the lesion segmentation, which prevents non-tumor related features to contribute to the classification of unhealthy samples.

TL;DR: FigureQA as discussed by the authors is a visual reasoning corpus of over one million question-answer pairs grounded in over 100,000 images, including line plots, dot-line plots, vertical and horizontal bar graphs, and pie charts.

TL;DR: In this article, a multi-inference trick is used to train all layers of the DBM simultaneously, using a novel training procedure called multi-prediction training, which can be interpreted as a single generative model trained to maximize a variational approximation to the generalized pseudolikelihood, or as a family of recurrent networks that share parameters and may be approximately averaged together.

TL;DR: Ethical principles and guidelines that surround machine learning and artificial intelligence are examined in a bid to clarify the role of language in the development of artificial intelligence.

TL;DR: This work abstracts the modeled dynamical system as a collection of autonomous but sparsely interacting sub-systems, which results in a class of models that are well suited for modeling the dynamics of systems that only offer local views into their state, along with corresponding spatial locations of those views.