<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Signals: Trajectory Sampling and Triage for Agentic Interactions · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2604.00356/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:27:26 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/2604.00356/feed.xml" rel="self" type="application/rss+xml" /><item><title>Signals: Trajectory Sampling and Triage for Agentic Interactions — HN submission</title><link>https://news.ycombinator.com/item?id=47645594</link><guid isPermaLink="false">urn:hackerpapers:submission:2604.00356:47645594</guid><pubDate>Sun, 05 Apr 2026 02:27:25 GMT</pubDate><description>&lt;p&gt;Show HN: Signals – finding the most informative agent traces without LLM judges&lt;/p&gt;&lt;p&gt;Hey HN Salman, Shuguang and Adil here from Katanemo Labs (a DigitalOcean company). Wanted to introduce our latest research on agentic systems called Signals. If you&amp;#x27;ve been building agents, you&amp;#x27;ve probably noticed that there are far too many agent traces/trajectories to review one by one, and using humans or extra LLM calls to inspect all of them gets expensive really fast. The paper proposes a lightweight way to compute structured “signals” from live agent interactions so you can surface the trajectories most worth looking at, without changing the agent’s online behavior. Computing Signals…&lt;/p&gt;&lt;p&gt;3 points; 0 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/2604.00356/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item></channel></rss>
