[D] NeurIPS 2020 Paper Reviews

If you're working on RL you should most definitively --not-- be submitting to NeurIPS/ICML/ICLR, unless it's focused on more old school topics like markov chains/MDPs, in which case you'll likely get responsible, senior reviewers and avoid the problem altogether. For RL you have IROS, ICRA, CoRL (which is new but growing), and although I haven't published in any of them, I've heard from colleagues who work on RL that they're all very functional and don't come even close to the mess that NeurIPS is right now. For vision and NLP you have the obvious ones. Both CVPR and ICCV take the reviewing process quite seriously compared to NeurIPS/ICML/ICLR. UAI is a great conference but it's quite niche, so it doesn't make a lot of sense to submit a work there unless it's about graphical models, causality, RL (more theoretical works) and other topics closer to stats.