Metraipolis › archive › dc_1cc27fb2df5f
Optimize the Harness, Not the Model: Fitting a Minimal Agent to a Self-Hostable LLM for Investigative Analysis — arXiv:2611.03434
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.03434
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T10:59:35.442Z
- SHA-256 of the archived text
92519cdbf42583c94031aa3aa533ad76dad340b410b1c956e555a5d58d869bcc- Stable URL of this page
https://metraipolis.com/documento/dc_1cc27fb2df5f
How to cite
Optimize the Harness, Not the Model: Fitting a Minimal Agent to a Self-Hostable LLM for Investigative Analysis — arXiv:2611.03434 — arXiv (en), https://arxiv.org/abs/2611.03434 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1cc27fb2df5f (Metraipolis archive dc_1cc27fb2df5f, sha256 92519cdbf42583c94031aa3aa533ad76dad340b410b1c956e555a5d58d869bcc)
BibTeX
@misc{dc1cc27fb2df5f,
title = {Optimize the Harness, Not the Model: Fitting a Minimal Agent to a Self-Hostable LLM for Investigative Analysis — arXiv:2611.03434},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.03434},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1cc27fb2df5f; sha256 92519cdbf42583c94031aa3aa533ad76dad340b410b1c956e555a5d58d869bcc},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03434. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Government intelligence teams need LLM agents that can analyze large, messy evidence collections, but security and cost constraints often rule out hosted frontier models and broad, hard-to-audit tool surfaces. We report CLEO: Constrained Loop for Environment Optimization, a method for fitting a minimal agent harness to a fixed, self-hostable LLM. Given an expert-validated dataset, CLEO admits only the tools, guardrails, memory policies, and loop components that improve accuracy while remaining general and auditable. On our held-out dataset, the CLEO harness reaches an 83.8% pass rate, within 1.2 points of the strongest out-of-the-box Claude Code SDK baseline, while being two orders of magnitude smaller. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.MA, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Multiagent Systems (cs.MA). Pagina del lavoro: https://arxiv.org/abs/2611.03434 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-03T10:59:22.317Z dalla pagina https://arxiv.org/list/cs.MA/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).