<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>WHALE: A Simple Recipe for Joint Harness-Weight Optimization · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2609.00196/</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/2609.00196/feed.xml" rel="self" type="application/rss+xml" /><item><title>WHALE: A Simple Recipe for Joint Harness-Weight Optimization — mentioned by htrp</title><link>https://news.ycombinator.com/item?id=50278985</link><guid isPermaLink="false">urn:hackerpapers:mention:2609.00196:50278985</guid><pubDate>Wed, 28 Oct 2026 18:50:32 GMT</pubDate><description>&lt;p&gt;&amp;gt;4. Although it&amp;#x27;s hard to 100% confirm this, there&amp;#x27;s reason to believe that Claude itself works better with Claude Code than with other harnesses. I thought this was pretty much well established that all of the labs have model - harness pairings for exactly this reason Whale Paper. https://arxiv.org/abs/2609.00196 EMNLP talk from like 2 weeks ago https://aclanthology.org/2026.findings-emnlp.2207/&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2609.00196/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.LG</category></item></channel></rss>
