I'm fascinated by the probabilistic numerics community realizing that their field was kickstarted in 1969, https://arxiv.org/pdf/1702.03673.pdf
Causality is this little field that I think could potentially really come into vogue in the coming years. This is a great set of links to get started with https://yanirseroussi.com/causal-inference-reading-list/
While deep models are still the hot thing, I massively respect a system that outputs risk scores like https://arxiv.org/abs/1610.00168
Just got back from #neurips if there were one highlight it is that Determental point processes are magical.
New dataset of hate speech out now, https://files.pushshift.io/gab/
Fitting ordinal data in Stan is basically built-in, https://gist.github.com/ericnovik/90e1c2377743135827d50e3313ae8968
Generalistic and moderated instance. All opinions are welcome, but hate speeches are prohibited. Users who don't respect rules will be silenced or suspended, depending on the violation severity.