arXiv cs.CL· Tianyu Wang·· 12 小时前AI 评分26
AMEND：用隔离评审与约束引导的多智能体文本去毒框架
AMEND: Constraint-Guided Multi-Agent Text Detoxification with Isolated Critics
AI 导读
AMEND 是一个推理时文本去毒框架，将去毒拆分为角色条件化候选生成、隔离的多维评审和约束优先选择，先强制满足语义与流畅度可行性再按效用排序。
正文 · 原文
Abstract: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%.
| Comments: | 10 pages, 2 figures, 4 tables |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.23183 [cs.CL] |
| (or arXiv:2610.23183v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.23183 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Tianyu Wang [view email]
[v1]
Fri, 16 Oct 2026 09:22:21 UTC (1,594 KB)
来源：arXiv cs.CL · arxiv.org