Metraipolis › archive › dc_ef0ebfa967fc

Reachable Global Optimization in AI Systems: How Global Is Global? — arXiv:2609.27855

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.27855
License (as archived): CC0-1.0
Archived: 2026-10-08T09:11:34.290Z
SHA-256 of the archived text: 1e367cf91bee2687af51b3bfecdc17cbb9b22dd2c5c8dfb866308f81c537eaad
Stable URL of this page: https://metraipolis.com/documento/dc_ef0ebfa967fc

How to cite

Reachable Global Optimization in AI Systems: How Global Is Global? — arXiv:2609.27855 — arXiv (en), https://arxiv.org/abs/2609.27855 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ef0ebfa967fc (Metraipolis archive dc_ef0ebfa967fc, sha256 1e367cf91bee2687af51b3bfecdc17cbb9b22dd2c5c8dfb866308f81c537eaad)

BibTeX

@misc{dcef0ebfa967fc,
  title = {Reachable Global Optimization in AI Systems: How Global Is Global? — arXiv:2609.27855},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.27855},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ef0ebfa967fc; sha256 1e367cf91bee2687af51b3bfecdc17cbb9b22dd2c5c8dfb866308f81c537eaad},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.27855. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

AI systems increasingly claim to optimize prompts, policies, architectures, plans, tool-use trajectories, reasoning traces, and test-time computation. This paper argues that such claims are underspecified unless they state the region actually reachable by the system that performed the optimization. We introduce Reachability-Induced Optimization (RIO), a model in which a generator, verifier, controller, memory, tools, and budget induce a reachable candidate region. The returned solution is therefore a best visited point, an approximate reachable optimum, or an exact global optimum only when additional certificates relate the reachable region to the full formal space. We prove reachable-optimality, false-globality, gap- decomposition, certificate, escape, pruning, and control-value results. The full benchmark record contains 66,150 executed trials over six known-optimum landscape families, seven control policies, 270 landscapes, and 35 runs per landscape-method. The online appendix includes raw trial records, aggregate tables, figures, benchmark code, validation scripts, and checksums. The results show that control can restrict, expand, or misdirect reachability, and that optimization quality, reachability quality, and control reliability must be reported separately.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.27855 — 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 2026-10-08T08:13:26.538Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).