Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Based on an initial read, this looks like a significant breakthrough to me.

I wonder if the techniques developed by the authors could make it feasible to take other piecewise linear/constant algorithms (which until now have been considered "non-differentiable" for practical purposes) and turn them into differentiable algorithms.

Think beyond sorting and ranking.



it’s not completely unprecedented because there were other ways of getting equivalent results before (the Sinkhorn based optimal transport approach cited, for one), which have been used for all kinds of interesting tasks. the contribution is that it does so more efficiently.


Agree. That's what I mean when I wrote "for practical purposes" above... although in hindsight I could have articulated it better. Thanks!


There's this work from the same team (posted the same day), more general problems with a different method

https://arxiv.org/abs/2002.08676




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: