Reviews: Thompson Sampling and Approximate Inference

This submission was diligently evaluated, and the main problem that was identified was clarity. An additional expert reviewer was asked to checked the proofs. I am suggesting acceptance, given that the authors put a serious effort into rewriting. The reviewers did a great job in identifying where the improvements are needed, please do take advantage of that. Furthermore, for the final version, please take a look at this paper: https://link.springer.com/chapter/10.1007%2F978-3-319-46379-7_22 .