It's great to have this all in one place - the change in notation across papers can be quite confusing!<p>I just wish there was something like this with code examples. It's so frustrating when papers say things like 'we omit biases' and so on because it makes it difficult to reproduce their implementation arghghg.