Skip to content
Trending storyDeveloping

Google and Northwestern propose Harness² for self-improving agent harnesses

1 article1 sourcesince Oct 15Last article 2h ago ·

Overview

AISummary of 1 article

Google and Northwestern researchers propose Harness², a framework that improves an AI agent's harness at test time without ground-truth verifiers, according to a paper summarized by Elvis Saravia on X. The method proposes parallel candidate edits, keeps the best combination, and carries lessons forward to a second harness.

The paper reports gains of up to 11.5 points on WorkBuddy-Bench, 4.5 points on JobBench, and 11.8 points on LAB, tested across GLM-5 and Gemini 3.5 Flash.

Written by AI from the articles below · updated Oct 15, 11:36 AM ET

Check the sources:

Latest developmentOct 15, 11:32 AM ET
Google paper proposes Harness² for recursive agent harness improvement on open-world tasks

Article timeline

The articles in this story. Times are ET.

Oct 15
  1. elvisX
    Google paper proposes Harness² for recursive agent harness improvement on open-world tasks

    AIGoogle and Northwestern researchers propose Harness², a framework that improves an agent's harness at test time without ground-truth verifiers. It proposes parallel candidate edits, keeps the best combination, and carries lessons forward to a second harness. The paper reports gains of up to 11.5 points on WorkBuddy-Bench, 4.5 on JobBench, and 11.8 points on LAB across GLM-5 and Gemini 3.5 Flash.

Heat trend

Not enough continuous observations to show a trend yet.