We’re excited to support today’s launch of Distill, a new kind of journal aimed at excellent communication of machine learning results (novel or existing).

Distill is a website and set of associated tools that make it easier for people to explain machine learning concepts using modern web technologies. For example, people have already used the platform to explore the subtle settings of the t-SNE algorithm⁠(opens in a new window), to demystify the checkerboard artifacts in synthetic images⁠(opens in a new window), and peek under the hood of recurrent neural networks that generate handwriting⁠(opens in a new window).