” On each island, a mutator agent selects a parent harness, studies its failure traces and lineage, then rewrites the whole harness rather than optimizing only a prompt or skill.
MILO co-evolves agent harnesses and the strategies that discover them
Across three agent benchmarks, adaptive multi-agent search produced stronger harnesses than eight existing harnesses and six automated search methods.
Big Tech
Prithwish Jana · Mononito Goswami · Hao Liu · Xinyu Li · Langlin Huang · Zhehui Huang · +6 more
Georgia Institute of Technology · AWS AI Labs · Carnegie Mellon University · Washington University in St.
Research Digest··3 min read
Thread:Agent Harness Optimization
Jana et al.
Why this paper
From AWS AI Labs and 3 others · Part of Agent Harness Optimization, now 82 papers
In one line
MILO co-evolves agent harnesses and search strategies via meta-evolutionary island orchestration, outperforming state-of-the-art methods.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ✓Reports numbers on named benchmarks (4 benchmarks)
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§