<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: tidewave</title><link>https://news.ycombinator.com/user?id=tidewave</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Wed, 16 Sep 2026 13:05:31 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=tidewave" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by tidewave in "Introducing System One Models and Jev"]]></title><description><![CDATA[
<p>Congrats on the release!<p>Finetuning a language model for decision classification (with probabilities) is already well-understood. What specifically changes in the training objective with RLCD? Are its benefits isolated from Jev’s new architecture/parallelism?</p>
]]></description><pubDate>Tue, 15 Sep 2026 20:42:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=49718646</link><dc:creator>tidewave</dc:creator><comments>https://news.ycombinator.com/item?id=49718646</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49718646</guid></item></channel></rss>