Metraipolis › archive › dc_da0806d0c0f7

When Are Aggregate Agent Traces Diagnosable? Traffic-Governed Interpretation and Calibrated Abstention — arXiv:2609.25806

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.25806
License (as archived)
CC0-1.0
Archived
2026-10-08T13:45:20.623Z
SHA-256 of the archived text
2f6001121fa55a98f6401469858b56eeaf8acc75131e01cfaacdb1297ada0a2f
Stable URL of this page
https://metraipolis.com/documento/dc_da0806d0c0f7

How to cite

When Are Aggregate Agent Traces Diagnosable? Traffic-Governed Interpretation and Calibrated Abstention — arXiv:2609.25806 — arXiv (en), https://arxiv.org/abs/2609.25806 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_da0806d0c0f7 (Metraipolis archive dc_da0806d0c0f7, sha256 2f6001121fa55a98f6401469858b56eeaf8acc75131e01cfaacdb1297ada0a2f)
BibTeX
@misc{dcda0806d0c0f7,
  title = {When Are Aggregate Agent Traces Diagnosable? Traffic-Governed Interpretation and Calibrated Abstention — arXiv:2609.25806},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.25806},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_da0806d0c0f7; sha256 2f6001121fa55a98f6401469858b56eeaf8acc75131e01cfaacdb1297ada0a2f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.25806. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Runtime traces can appear transparent, but a closed-loop policy determines which states are visited and which failures become visible. We study a simulated hotel-pricing agent mapping time, inventory, and market state to discrete price actions under varying demand regimes. A fault may leave no aggregate trace when the policy rarely visits affected cells. We treat entry into aggregate-only fault interpretation as a diagnosability decision preceding scoring or localization. A reference-map gate requires repeated clean-policy support; a matched runtime gate then requires joint support in clean and current streams. Signal analysis occurs only after both pass. We calibrate false admission on a disjoint clean stream at the physical-component level and model detection by affected clean traffic rather than nominal cell coverage. In a frozen one-shot heldout, 55/72 (76.4%) regime-component units were reference-admitted, representing 20 physical components; 54/55 passed matched runtime admission, while the rejected unit abstained. Stable false admission was 0/20, with a one-sided exact 95% upper bound of 0.1391, meeting the frozen 0.20 criterion. Across 540 repeated unit-arm rows nested in those 20 clusters, affected clean traffic reduced negative log likelihood by 29.3% relative to cell coverage, a gain of 0.1264 nats per row (cluster-bootstrap 95% interval [0.0593, 0.1918]). Adding mask family and its interaction improved log loss by 0.0015 nats per row (one-sided upper bound 0.0066), below the frozen 0.01 practical-sufficiency margin. A development audit found that exact minimum hitting set and greedy selection chose identical supports in 12/12 scenarios because singleton evidence had resolved the conflicts. The result is a bounded rule for interpreting aggregate agent behavior: first establish exposure, then score change, and abstain when the trace cannot support the claim.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-23 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.25806 — 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 2026-10-08T13:18:38.125Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-23?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).