Metraipolis › archive › dc_4d8ac69fbd33
TUPLE-Bench: Evaluating and Mitigating Privacy-Utility Trade-offs in Typed Decision Models — arXiv:2610.23668
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.23668
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T08:43:19.956Z
- SHA-256 of the archived text
5e46ba1a16a5aee492f5da139ee770007d8afc4053c439e6d21a7b53d8bb4b58- Stable URL of this page
https://metraipolis.com/documento/dc_4d8ac69fbd33
How to cite
TUPLE-Bench: Evaluating and Mitigating Privacy-Utility Trade-offs in Typed Decision Models — arXiv:2610.23668 — arXiv (en), https://arxiv.org/abs/2610.23668 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4d8ac69fbd33 (Metraipolis archive dc_4d8ac69fbd33, sha256 5e46ba1a16a5aee492f5da139ee770007d8afc4053c439e6d21a7b53d8bb4b58)
BibTeX
@misc{dc4d8ac69fbd33,
title = {TUPLE-Bench: Evaluating and Mitigating Privacy-Utility Trade-offs in Typed Decision Models — arXiv:2610.23668},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.23668},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4d8ac69fbd33; sha256 5e46ba1a16a5aee492f5da139ee770007d8afc4053c439e6d21a7b53d8bb4b58},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.23668. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Typed decision models often return more than a selected category. They also expose probabilities for each candidate and, sometimes, a separate confidence score. These fields support downstream decisions but may reveal information beyond a recipient's authorization. We introduce TUPLE-Bench (Typed decision models' Utility and Privacy Leakage Evaluation Benchmark), a benchmark of paired synthetic cases in which recipients may learn a coarse operational category but not its fine-grained cause. Across nine models, instructions to withhold the cause fail to protect it: 91.88%-99.93% of responses reveal the cause through direct selection or residual scores. Even when causes are excluded from the candidate set, full outputs enable 62.85%-75.35% recovery on held-out cases, compared with 51.04%-57.64% using only the selected category (50% chance). Full outputs also reduce mean downstream cost in all 27 model-task comparisons. However, confidence fields increase leakage without consistently reducing cost. We further introduce workflow-constrained privacy quantization (WCPQ), which reduces secret recovery while preserving the behavior and cost of specified downstream workflows. Our findings highlight the need to audit decision interfaces field by field and workflow by workflow. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.23668 — 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 2026-10-19T08:39:10.163Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).