<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>AdaSD: Adaptive Speculative Decoding for Efficient Language Model Inference · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2512.11280/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/2512.11280/feed.xml" rel="self" type="application/rss+xml" /><item><title>AdaSD: Adaptive Speculative Decoding for Efficient Language Model Inference — mentioned by hypfer</title><link>https://news.ycombinator.com/item?id=49302745</link><guid isPermaLink="false">urn:hackerpapers:mention:2512.11280:49302745</guid><pubDate>Fri, 14 Aug 2026 18:27:45 GMT</pubDate><description>&lt;p&gt;I wonder if we could take a page out of the Solar power book and do MPPT but for draft count. Constantly sweep through possible values to find the best result for the current conditions. Though it might turn out that that doesn&amp;#x27;t offer any benefits. Has it been tried yet? ___ EDIT: LLM pointed me towards this thing I&amp;#x27;m not going to read https://arxiv.org/abs/2512.11280 &amp;quot;AdaSD: Adaptive Speculative Decoding for Efficient Language Model Inference&amp;quot; Some higher level nerd please go have a look at that. I have important meetings and golf classes to attend.&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2512.11280/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item></channel></rss>
