Metraipolis › archive › dc_fff8bed678ff

The "10th Juror": Open-Set Standpoint Screening for Bureaucratic Bias Detection — arXiv:2610.11136

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.11136
License (as archived): CC0-1.0
Archived: 2026-10-09T06:04:08.036Z
SHA-256 of the archived text: 82c1ec94c5870bb2e54a8c48f059513462f14db4ef32b2f3b9c9a2337f438708
Stable URL of this page: https://metraipolis.com/documento/dc_fff8bed678ff

How to cite

The "10th Juror": Open-Set Standpoint Screening for Bureaucratic Bias Detection — arXiv:2610.11136 — arXiv (en), https://arxiv.org/abs/2610.11136 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fff8bed678ff (Metraipolis archive dc_fff8bed678ff, sha256 82c1ec94c5870bb2e54a8c48f059513462f14db4ef32b2f3b9c9a2337f438708)

BibTeX

@misc{dcfff8bed678ff,
  title = {The "10th Juror": Open-Set Standpoint Screening for Bureaucratic Bias Detection — arXiv:2610.11136},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11136},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fff8bed678ff; sha256 82c1ec94c5870bb2e54a8c48f059513462f14db4ef32b2f3b9c9a2337f438708},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11136. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Presupposing the boundaries of bias is itself a form of bias. We study closed-loop bias governance for Dutch government documents, where a system must detect biased language, ground decisions in legal and contextual evidence, rewrite problematic sentences when intervention is warranted, and verify that the rewrite mitigates harm without distorting meaning. Existing methods face three challenges: (i) discriminative classifiers capture surface regularities but lack normative grounding; (ii) zero-shot LLMs often adopt generic viewpoints and over-flag ambiguous administrative language; and (iii) fixed taxonomies inherit the Closed-World Assumption, missing emerging local targets. We propose MARS-Gov, a standpoint-aware multi-agent framework that combines legal retrieval, open-set target screening, specialized jurors, conservative routing, and rewrite verification. When screening finds an uncovered group, MARS-Gov instantiates a dynamic "10th juror" to deliberate outside the fixed panel. On DGDB, MARS-Gov sets a new SOTA with 0.880 F1, outperforming the strongest zero-shot LLM detector by 20.2 points (29.8% relative) and the best supervised Dutch encoder by 6.8 points, while reducing unnecessary interventions to 2.5%. Leave-One-Category-Out (LOCO) evaluation recovers held-out categories with 85.1% Correct@1 and 93.8% Correct@3.

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/2610.11136 — 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 2026-10-09T05:47:15.117Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).