https://hackerpaper.kaipfister.de/p/2111.09266/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMThttps://news.ycombinator.com/item?id=38152855urn:hackerpapers:mention:2111.09266:38152855Sun, 05 Nov 2023 16:51:38 GMT<p>You can use the straight through operator! During the forward pass you sample a discrete outcome given your NN weights to get an error for backprop. During the backward pass you directly propogate through the weights. This GradTree paper[1] does a good job covering how to do discrete gradient-based optimization (i.e. NNs w/ discrete representations). Another option is to use a GFlowNet[2]. Then you have a NN policy that takes discrete actions like you&#x27;re playing an RL game. You&#x27;re not back-propogating through something that isn&#x27;t continuous, but you&#x27;re utilizing a NN to make informed…</p><p><a href="https://hackerpaper.kaipfister.de/p/2111.09266/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.LG