Metraipolis › archive › dc_2541e7e8ec06

AMEND: Constraint-Guided Multi-Agent Text Detoxification with Isolated Critics — arXiv:2610.23183

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.23183
License (as archived)
CC0-1.0
Archived
2026-10-19T08:43:46.134Z
SHA-256 of the archived text
1dce912fe7a83e7daa9ca3f36a1e85b49112f22f3b0a361ca42637f7c43c1711
Stable URL of this page
https://metraipolis.com/documento/dc_2541e7e8ec06

How to cite

AMEND: Constraint-Guided Multi-Agent Text Detoxification with Isolated Critics — arXiv:2610.23183 — arXiv (en), https://arxiv.org/abs/2610.23183 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2541e7e8ec06 (Metraipolis archive dc_2541e7e8ec06, sha256 1dce912fe7a83e7daa9ca3f36a1e85b49112f22f3b0a361ca42637f7c43c1711)
BibTeX
@misc{dc2541e7e8ec06,
  title = {AMEND: Constraint-Guided Multi-Agent Text Detoxification with Isolated Critics — arXiv:2610.23183},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.23183},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2541e7e8ec06; sha256 1dce912fe7a83e7daa9ca3f36a1e85b49112f22f3b0a361ca42637f7c43c1711},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.23183. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Text detoxification requires balancing toxicity removal, semantic preservation, and fluency. Existing methods typically optimize an unconstrained scalar objective---enabling safety scores to mask semantic loss---or rely on open-ended iterative debate prone to semantic drift and high computational cost. We propose AMEND, an inference-time framework that decouples detoxification into role-conditioned proposal generation, isolated multidimensional criticism, and constraint-first selection. AMEND evaluates diverse candidates with role-blind critics and enforces strict semantic and fluency feasibility before ranking by utility. To maximize efficiency, AMEND Lite restricts subsequent search to at most one targeted repair round only when residual toxic spans remain. On ParaDetox, one-round selection over diverse roles improves the joint evaluation metric by $+0.0508$ over the best single role. On a frozen confirmation test set, AMEND Lite achieves a joint score of $0.6597 \pm 0.0062$, outperforming unconstrained multi-round loops by $+0.0489$ ($p < 0.05$) while reducing inference requests by 28.4%.

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/2610.23183 — 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 2026-10-19T08:39:24.860Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).