Metraipolis › archive › dc_a5c3e4e80f45
JevOut: Natural Context Can Flip Decision Models — arXiv:2609.30243
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.30243
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T04:50:07.168Z
- SHA-256 of the archived text
aff8c3fa835d63aad1f859eb251a2a8dd3cda643280d7193661fbc4f61d71104- Stable URL of this page
https://metraipolis.com/documento/dc_a5c3e4e80f45
How to cite
JevOut: Natural Context Can Flip Decision Models — arXiv:2609.30243 — arXiv (en), https://arxiv.org/abs/2609.30243 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a5c3e4e80f45 (Metraipolis archive dc_a5c3e4e80f45, sha256 aff8c3fa835d63aad1f859eb251a2a8dd3cda643280d7193661fbc4f61d71104)
BibTeX
@misc{dca5c3e4e80f45,
title = {JevOut: Natural Context Can Flip Decision Models — arXiv:2609.30243},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.30243},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a5c3e4e80f45; sha256 aff8c3fa835d63aad1f859eb251a2a8dd3cda643280d7193661fbc4f61d71104},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.30243. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Dedicated decision models such as Jev map unstructured language to probability distributions over finite choices, allowing their outputs to directly route requests, select tools, and trigger actions. Yet real-world inputs rarely arrive in isolation: they come with background details and surrounding context. We find that short additions that fit naturally into this context can nevertheless redirect an otherwise correct decision, even when the correct answer remains unchanged. To study this behavior, we fix a wrong target option for each initially correct item and use the model's option probabilities to refine fluent context additions while preserving the source, question, choices, and gold answer. Within 64 accepted target evaluations, the optimizer identifies contexts that redirect Jev on 312 of 508 initially correct decisions (61.4%); in 229 cases, Jev assigns at least 0.7 probability to the fixed wrong option. Across seven datasets, three additional decision systems show targeted flip rates of 64.9%-73.2% on decisions they initially answer correctly. Taken together, these results expose a pronounced fragility in current decision models: short, ordinary-looking context can shift a correct choice to a high-confidence wrong one. Because these models turn language directly into downstream choices, this sensitivity raises concerns about treating their probability outputs as reliable decision interfaces. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-25 in cs.CL). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2609.30243 — 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 2026-10-08T04:27:20.140Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).