Metraipolis › archive › dc_ffcbe4ccd2a5

DecisionRobust: Benchmarking Robustness for JEV Style Decision Models Across Text and Vision — arXiv:2610.19118

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.19118
License (as archived)
CC0-1.0
Archived
2026-10-15T20:23:30.269Z
SHA-256 of the archived text
dcfbc7bbd629a58631fad01a0bdcb4a7f2ae3fd665bb3ccd57f82b4202bc33c8
Stable URL of this page
https://metraipolis.com/documento/dc_ffcbe4ccd2a5

How to cite

DecisionRobust: Benchmarking Robustness for JEV Style Decision Models Across Text and Vision — arXiv:2610.19118 — arXiv (en), https://arxiv.org/abs/2610.19118 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ffcbe4ccd2a5 (Metraipolis archive dc_ffcbe4ccd2a5, sha256 dcfbc7bbd629a58631fad01a0bdcb4a7f2ae3fd665bb3ccd57f82b4202bc33c8)
BibTeX
@misc{dcffcbe4ccd2a5,
  title = {DecisionRobust: Benchmarking Robustness for JEV Style Decision Models Across Text and Vision — arXiv:2610.19118},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.19118},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ffcbe4ccd2a5; sha256 dcfbc7bbd629a58631fad01a0bdcb4a7f2ae3fd665bb3ccd57f82b4202bc33c8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19118. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Emerging decision models such as JEV provide typed interfaces that map task instructions and observations directly to choices, judgments, and scores. As these models are increasingly used for automated decision-making, understanding their reliability under adversarial manipulation becomes critical. We introduce DecisionRobust, a benchmark for systematically evaluating the robustness of decision models across textual and visual inputs. Our benchmark covers practical decision-making scenarios and a diverse suite of attacks, including semantics-preserving text perturbations, image perturbations, and prompt injection in both modalities. We curate evaluation sets through external validity checks and cross-model candidate selection. We evaluate 30 model configurations on text tasks and 17 on vision tasks, including 24 and 12 held-out models, respectively, that are excluded from candidate selection. Our experiments reveal several key findings. First, attacks expose vulnerabilities beyond the models used for construction and selection, although optimized image perturbations transfer weakly to many targets even with multiple surrogates. Second, relative model robustness varies across tasks and attack types, and resistance to input perturbations does not necessarily imply resistance to prompt injection. Third, label-consistency checks substantially reduce measured attack success across text and image settings, highlighting the importance of validating apparent failures. The benchmark is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.19118 — 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 2026-10-15T19:36:18.164Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).