Stephen Downes

This website is daunting, to say the least. It's a branch of Papers With Code, a challenge in its own right. The idea here is to collect papers that are accompanied with computer code (which is freely accessible on GitHub). But more, the papers are grouped into categories, and each category is subgrouped into tasks, and then the papers are sorted by their approach to that task, and 'state-of-the-art' is defined as the most successful approach currently. So basically, what you have are problems, proposed solutions, and benchmarks, all out in the open. This is what open science looks like, and I'd love to see what an application of the same idea in the domains of learning, inference and discovery would look like. Via Reddit.