AI Daybook

From the archiveNo. 2

Open Source & Tools

Harness-1, a 20B retrieval subagent trained with reinforcement learning, reaches 0.730 curated recall

Meet Harness-1: A 20B Retrieval Subagent Trained With Reinforcement Learning Inside a Stateful Search Harness on gpt-oss-20b

MarkTechPost

Summary

UIUC and Chroma released Harness-1, a 20-billion-parameter retrieval subagent trained with reinforcement learning inside a stateful search harness. It achieves 0.730 average curated recall across eight benchmarks, beating the next open subagent by 11.4 points and trailing only Opus-4.6.

Read the original

#rag#harness-1#retrieval agent

142 more stories in this issueNo. 2 →Tue, Jun 9, 2026