Metraipolis › archive › dc_ee7f15e51907
Wavering Oracles: Selective Updating and Correlated Failures in LLMs and Their Implications for Scientific Workflows — arXiv:2609.11428
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.11428
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T05:45:44.818Z
- SHA-256 of the archived text
a707ae70ba4e331fece599737a8d9263916832043961d2e0c3d05e2ae5b51375- Stable URL of this page
https://metraipolis.com/documento/dc_ee7f15e51907
How to cite
Wavering Oracles: Selective Updating and Correlated Failures in LLMs and Their Implications for Scientific Workflows — arXiv:2609.11428 — arXiv (en), https://arxiv.org/abs/2609.11428 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ee7f15e51907 (Metraipolis archive dc_ee7f15e51907, sha256 a707ae70ba4e331fece599737a8d9263916832043961d2e0c3d05e2ae5b51375)
BibTeX
@misc{dcee7f15e51907,
title = {Wavering Oracles: Selective Updating and Correlated Failures in LLMs and Their Implications for Scientific Workflows — arXiv:2609.11428},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.11428},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ee7f15e51907; sha256 a707ae70ba4e331fece599737a8d9263916832043961d2e0c3d05e2ae5b51375},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.11428. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Scientific workflows increasingly use repeated queries, multiple models, and interacting agents. Reliability therefore depends on whether models preserve correct conclusions, accept valid corrections, and contribute errors that a selector can distinguish. Using SycoBench- 600 as a controlled measurement substrate, we evaluate these requirements through selective updating, defined by resistance to misleading suggestions and uptake of correct suggestions. The study covers ten models and 17,055 trajectories. Published models span 13.4 to 71.6 percentage points in selectivity. Under identical local evaluation, Qwen3-4B is selectively adaptive at 45.6 points, Gemma3-4B is destabilized at minus 14.1 points, and SmolLM3-3B follows both correct and wrong explicit suggestions, producing zero selectivity. Matched interventions identify model specific responses to doubt, authority, and explicit advice. Among seven published models, the best reaches 95.3 percent accuracy, plurality reaches 88.6 percent, and the oracle ceiling is 99.8 percent. Mean error correlation of 0.285 reduces seven models to an effective independent count of 2.58. A leave-one-stem-family-out reliability selector reaches 96.2 percent, recovering 67.7 percent of the plurality-to-oracle gap. These results establish selective updating, error diversity, and calibrated adjudication as jointly measurable design targets for multi-model scientific workflows. Tema degli esploratori: ricerca e scienza aperta (retro-catalogo arXiv: annunci del 2026-09-11 in cs.DL). Categorie arXiv: Digital Libraries (cs.DL). Pagina del lavoro: https://arxiv.org/abs/2609.11428 — 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 2026-10-16T05:30:07.931Z dalla pagina https://arxiv.org/catchup/cs.DL/2026-09-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).