<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: hjun1052</title><link>https://news.ycombinator.com/user?id=hjun1052</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Mon, 12 Oct 2026 01:27:32 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=hjun1052" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by hjun1052 in "Show HN: Happy Hangul Day! An RNN for Generating Korean Handwriting Strokes"]]></title><description><![CDATA[
<p>Wow, what a cool project! It struggles a bit with "쀍" tho... (but so does my phone’s Korean font lol)</p>
]]></description><pubDate>Sun, 11 Oct 2026 10:23:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=50041597</link><dc:creator>hjun1052</dc:creator><comments>https://news.ycombinator.com/item?id=50041597</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50041597</guid></item><item><title><![CDATA[New comment by hjun1052 in "Jeeves. Reasoning improves Jev-like decision models"]]></title><description><![CDATA[
<p>If the model does autoregressive reasoning before the decision, doesn't that give up much of what a Jev-style model buys you (a single forward pass, cheap calibrated probabilities)? Or is the point mainly to keep the typed output and probability interface while getting better accuracy on harder cases?</p>
]]></description><pubDate>Tue, 29 Sep 2026 11:50:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=49891626</link><dc:creator>hjun1052</dc:creator><comments>https://news.ycombinator.com/item?id=49891626</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49891626</guid></item></channel></rss>