Metraipolis › archive › dc_fe158a8636e4
SEPO: Evidence-Grounded Prompt Optimization via Structural Editing — arXiv:2608.28067
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.28067
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T10:47:27.219Z
- SHA-256 of the archived text
80ef6068e40e24df2308190d01f74fc2ac874e813456fa35d99860da314988b4- Stable URL of this page
https://metraipolis.com/documento/dc_fe158a8636e4
How to cite
SEPO: Evidence-Grounded Prompt Optimization via Structural Editing — arXiv:2608.28067 — arXiv (en), https://arxiv.org/abs/2608.28067 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fe158a8636e4 (Metraipolis archive dc_fe158a8636e4, sha256 80ef6068e40e24df2308190d01f74fc2ac874e813456fa35d99860da314988b4)
BibTeX
@misc{dcfe158a8636e4,
title = {SEPO: Evidence-Grounded Prompt Optimization via Structural Editing — arXiv:2608.28067},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.28067},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fe158a8636e4; sha256 80ef6068e40e24df2308190d01f74fc2ac874e813456fa35d99860da314988b4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28067. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Existing API-only prompt optimisers are often described as interpretable, but in practice, this usually means only post-hoc inspectability: each iteration still rewrites the prompt as one opaque string, leaving a trace of full-prompt diffs rather than localisable, machine-readable edits. This paper introduces SEPO (Structural, Evidence-grounded Prompt Optimization), a multi-trajectory prompt optimiser centred on edit-effect lineage feedback. Rather than treating each iteration as an isolated whole-prompt rewrite, SEPO locally edits stable, typed units in a two-layer prompt schema, links the target and realised structural operations of each edit to the examples it newly fixes or breaks, and carries this edit-effect record forward to guide later architect calls on the same search branch. This makes prompt optimisation addressable, attributable, and actionable. Across a 14-task held-out suite, SEPO improves over the strongest baseline, GEPA, by 3.1 pp on Llama-3.1-8B-Instruct and 2.2 pp on Qwen3-8B, reaching 61.9% and 73.3% macro accuracy. SEPO also lies on both the optimisation-time and test-time Pareto frontiers, spending 2.9M optimisation tokens versus 4.1M for GEPA and producing prompts over 5x shorter. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-31 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.28067 — 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 2026-10-17T10:33:42.391Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-31?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).