When first deploying an anomaly detection system, e.g., to detect out-of-scope queries in chatbots, there are no observed data, making data-driven approaches ineffective. Zero-shot anomaly detection methods offer a solution to such "cold-start" cases, but unfortunately they are often not accurate enough. This paper studies the realistic but underexplored cold-start setting where an anomaly detection model is initialized using zero-shot guidance, but subsequently receives a small number of contaminated observations (namely, that may include anomalies). The goal is to make efficient use of both the zero-shot guidance and the observations. We propose ColdFusion, a method that effectively adapts the zero-shot anomaly detector to contaminated observations. To support future development of this new setting, we propose an evaluation suite consisting of evaluation protocols and metrics.

翻译：首次部署异常检测系统时（例如用于检测聊天机器人中超出范围的查询），由于缺乏观测数据，数据驱动方法往往失效。零样本异常检测方法为此类"冷启动"场景提供了解决方案，但其准确性通常不足。本文研究了一种现实存在但尚未充分探索的冷启动场景：异常检测模型通过零样本引导初始化后，会接收到少量受污染的观测数据（即可能包含异常值）。研究目标在于高效利用零样本引导与观测数据。我们提出ColdFusion方法，能够有效使零样本异常检测器适应受污染的观测数据。为支持这一新场景的未来发展，我们构建了包含评估协议与指标的评价体系。