
  <rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
      <title>Will Hampson &lt;-- Not a Bot</title>
      <link>https://whamp.github.io/blog</link>
      <description>Meandering Curiosity</description>
      <language>en-us</language>
      <managingEditor> (Will)</managingEditor>
      <webMaster> (Will)</webMaster>
      <lastBuildDate>Fri, 01 May 2026 19:00:00 GMT</lastBuildDate>
      <atom:link href="https://whamp.github.io/tags/qwen/feed.xml" rel="self" type="application/rss+xml"/>
      
  <item>
    <guid>https://whamp.github.io/blog/qwen-3-6-27b-livecodebench-reasoning-budget</guid>
    <title>I capped Qwen3.6&#39;s reasoning tokens. 8192 matched free thinking.</title>
    <link>https://whamp.github.io/blog/qwen-3-6-27b-livecodebench-reasoning-budget</link>
    <description>A LiveCodeBench experiment with Qwen3.6-27B, vLLM reasoning budgets, grammar-constrained reasoning, and the harness work needed to compare accuracy, token cost, and generation time.</description>
    <pubDate>Fri, 01 May 2026 19:00:00 GMT</pubDate>
    <author> (Will)</author>
    <category>qwen</category><category>livecodebench</category><category>reasoning</category><category>benchmarks</category>
  </item>

    </channel>
  </rss>
