LatentMoE: Toward Optimal Accuracy per FLOP and Parameter in Mixture of Experts · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2601.18089/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTLatentMoE: Toward Optimal Accuracy per FLOP and Parameter in Mixture of Experts — mentioned by buildbothttps://news.ycombinator.com/item?id=48935781urn:hackerpapers:mention:2601.18089:48935781Thu, 16 Jul 2026 15:19:52 GMT<p>Amazing to see an open source model already nearing the benchmarks of Fable and GPT 5.6 Sol! Also very cool to see LatentMoE being picked up by more models (https://arxiv.org/abs/2601.18089)</p><p><a href="https://hackerpaper.kaipfister.de/p/2601.18089/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.LG