Metraipolis › archive › dc_181eebc49c30

Explanatory Engagement Under Rare Anomalous Failure: Asymptotic Rarity in Model Behavior (or: The Asymptotic AI) — arXiv:2608.13063

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.13063
License (as archived)
CC0-1.0
Archived
2026-10-19T04:17:30.333Z
SHA-256 of the archived text
caac1acc4f2eb7ba32119d9eccc4e5b7045cec2bec00ff785a642d92fbf98f40
Stable URL of this page
https://metraipolis.com/documento/dc_181eebc49c30

How to cite

Explanatory Engagement Under Rare Anomalous Failure: Asymptotic Rarity in Model Behavior (or: The Asymptotic AI) — arXiv:2608.13063 — arXiv (en), https://arxiv.org/abs/2608.13063 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_181eebc49c30 (Metraipolis archive dc_181eebc49c30, sha256 caac1acc4f2eb7ba32119d9eccc4e5b7045cec2bec00ff785a642d92fbf98f40)
BibTeX
@misc{dc181eebc49c30,
  title = {Explanatory Engagement Under Rare Anomalous Failure: Asymptotic Rarity in Model Behavior (or: The Asymptotic AI) — arXiv:2608.13063},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.13063},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_181eebc49c30; sha256 caac1acc4f2eb7ba32119d9eccc4e5b7045cec2bec00ff785a642d92fbf98f40},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.13063. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Prior work on LLM behavior under anomalous conditions asks whether a model notices anomalies. We ask a narrower question: once a model sits in a workflow with a low, controllable failure rate, does its explanatory engagement - length, specificity, self-reported confidence - change as failure grows asymptotically rarer? We built a local, zero-cost harness on three open-weight models (qwen3:8b, llama3.1:8b, mistral:7b) running a repeated tool-call task where one call fails at probability p, swept across eight rates from 0.2 to 0.0001, under five elicitation conditions from immediate prompting to none. We hypothesized a rise in engagement as failures grew rarer, then a collapse near a detectability threshold. Pooled across conditions this appeared false: length fell in a flat, monotonic pattern. Splitting by condition overturned that. Under immediate_forced, where the model must explain every failure instantly, the predicted rise is confirmed but followed by a plateau, not a collapse: length peaks at 28.4 words at p=0.05, settles to 17.4-19.0 words at the rarest rates, and confidence rises unevenly from about 53% to the 70s-90s. Under grouped_runs, explanation batched to run-end, no collapse appears. Under passive_unprompted, aggregate magnitude is a floor artifact, but a recovered logging gap revealed real, model-specific self-monitoring: llama3.1:8b volunteers structured confidence reports unprompted, sometimes eroding its own confidence as trials accumulate; the other two do so only once, as boilerplate. Elicitation structure is a first-class moderator of collapse observability. A companion guaranteed-failure run (72 cells, backfilling rates where random sampling gave zero real failures) shows models differ in whether they recognize an anomaly, distinct from engagement once recognized. Limitation: discrete rate points cannot capture behavior between them, a direction for future work.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.13063 — 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 2026-10-19T04:13:06.573Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).