Meta · California · AI Agent · Crypto Briefing
Meta, Duke and UC Davis researchers unveil self-improving branches for agent harness optimization
Compiled by KHAO Editorial — aggregated from 1 source. See llms.txt for citation guidance.
◌ Single Source
Meta official logo (public domain, Wikimedia Commons), CryptoBriefing brand treatment.
Key facts
- Their preprint, titled “Mixture of Self-Improving Branches for Agent Harness Optimization,” reports a 34.8% relative improvement on Olympiad-level math reasoning
- Meta-Harness, released March 30, 2026, had previously outperformed traditional methods across various benchmarks
- The author list includes Haoyu Dong, affiliated with Meta and Duke University, and Zihao Lin, affiliated with Meta and UC Davis
- A jump from 46.0% to 62.0% on hard math, achieved without touching model weights, suggests meaningful improvements can come from engineering the environment a model operates
Summary
A new preprint splits harness search into specialized branches and routes each task to the best one, beating the earlier Meta-Harness system on math, terminal and coding benchmarks. Researchers from Meta, Duke University and the University of California, Davis have a new idea for making AI agents better. Their preprint, titled “Mixture of Self-Improving Branches for Agent Harness Optimization,” reports a 34.8% relative improvement on Olympiad-level math reasoning. More formally, an agent harness is the code framework wrapped around a large language model.