https://hackerpaper.kaipfister.de/p/2410.13780/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMThttps://news.ycombinator.com/item?id=47920133urn:hackerpapers:mention:2410.13780:47920133Mon, 27 Apr 2026 11:17:29 GMT<p>There are also more papers on similar themes. For example, TurboQuant makes use of QJL (quantized Johnson Lindenstrauss transformations). One of the first papers to characterize the QJL and in fact the rate distortion tradeoff for quantized matrix multiplication in general is &quot;Optimal Quantization for Matrix Multiplication&quot; (https://arxiv.org/abs/2410.13780) by Ordentlich and Polyanskiy. There is also a more accessible survey paper around quantized matrix multiplication called &quot;High-Rate Quantized Matrix Multiplication: Theory and Practice&quot; (https://arxiv.org/abs/2601.17187), by the same…</p><p><a href="https://hackerpaper.kaipfister.de/p/2410.13780/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.IThttps://news.ycombinator.com/item?id=41946734urn:hackerpapers:mention:2410.13780:41946734Fri, 25 Oct 2024 16:19:02 GMT<p>You might like an information-theoretic take on SpinQuant and the likes [1]. tl;dr: round((2*R)*x) is not a great idea for an R-bit quantization. [1] https://arxiv.org/abs/2410.13780</p><p><a href="https://hackerpaper.kaipfister.de/p/2410.13780/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.IT