<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"><channel><title>The Raw Logs — #benchmarks</title><link>https://therawlogs.com/tag/benchmarks</link><description>Top takes tagged #benchmarks.</description><language>en</language><lastBuildDate>Tue, 08 Sep 2026 16:37:33 GMT</lastBuildDate>    <item><title>#benchmarks — Aikido: specialized vuln models beat general LLMs at half the cost</title><link>https://therawlogs.com/take/14-4</link><guid>https://therawlogs.com/take/14-4</guid><pubDate>Sun, 30 Aug 2026 16:15:26 GMT</pubDate><description><![CDATA[Aikido: specialized vuln models beat general LLMs at half the cost: Aikido evaluated cyber AI with 11.7B tokens and proved specialized vulnerability models find more code bugs at 50% lower cost than general closed models. General models have high false positives. Specialized vulnerability scanners secure code faster for less compute. [#benchmarks]]]></description></item></channel></rss>