https://hackerpaper.kaipfister.de/p/1606.08415/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMThttps://news.ycombinator.com/item?id=38449438urn:hackerpapers:mention:1606.08415:38449438Tue, 28 Nov 2023 18:39:13 GMT<p>&gt; The field of ML research is moving so fast that people don’t even take time anymore to explain the design decisions behind their architectures. Has it ever? I&#x27;ve read a lot of &quot;old&quot; papers and idk if there&#x27;s ever been a strong theoretical framework. I&#x27;d say it is strongest today but not in main papers and those papers are getting rejected because of lack of experiments (wtf is going on with reviewing and why are our solutions &quot;authors have to show reviewers used LLMs&quot;[-1] not &quot;reviewers have to provide meaningful review&quot;? What a joke!). Here&#x27;s a more recent Universal Approximation paper[0].…</p><p><a href="https://hackerpaper.kaipfister.de/p/1606.08415/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.LGhttps://news.ycombinator.com/item?id=21546277urn:hackerpapers:mention:1606.08415:21546277Fri, 15 Nov 2019 16:39:05 GMT<p>Most of the recent research is moving to GELU (Gaussian Error Linear Units) activation functions: https://arxiv.org/pdf/1606.08415.pdf</p><p><a href="https://hackerpaper.kaipfister.de/p/1606.08415/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.LG