ASIRF：免重训练的敏感信息脱敏智能体框架
ASIRF: An Agentic Framework for Context-Dependent Sensitive Information Redaction
脱敏工具换个领域就得重训练？这篇论文用智能体检索领域定义，几十条规则就能超越 OpenAI Privacy Filter 的 85% 组合。
arXiv 论文提出 ASIRF（Agentic Sensitive Information Redaction Framework），在推理时从知识库检索领域相关的敏感信息定义，无需重新训练即可适配新领域。论文实现了三次调用的多智能体流水线和单智能体两种架构，在 10 个小型开源模型、8 个数据集上评测，并以 OpenAI Privacy Filter（OPF）作为训练分类器基线。每个领域只需几十条专家编写的定义、零训练数据，ASIRF 在 80 个模型-领域组合中的 68 个（85%）召回率超过 OPF，落后场景主要集中于 OPF 的训练分布内领域。
ASIRF: An Agentic Framework for Context-Dependent Sensitive Information Redaction
Sensitive information is defined by domain and intent, not a universal category, yet redaction systems such as privacy filters and named-entity recognizers fix a taxonomy at training time, requiring retraining for each new domain. We introduce ASIRF (Agentic Sensitive Information Redaction Framework), which retrieves domain-specific definitions based on the input's domain from a flexible knowledge base at inference time, needing no retraining to adapt. Two architectures, a three-call multi-agent pipeline and a single-agent variant, are evaluated across ten small open-weight models and eight datasets, including out-of-distribution fictional domains, against the OpenAI Privacy Filter (OPF) as a trained-classifier baseline. With only a few dozen expert-authored definitions per domain and no training data, ASIRF's recall exceeds OPF's in 68 of 80 model-domain combinations (85 percent), by at least one of the two architectures, with shortfalls confined mostly to OPF's training-distribution domains.
