AnonyRAG is a question-answering dataset for entity anonymization tasks released by Tencent Youtu Lab, Monash University and Hong Kong Polytechnic University in 2025. The related paper results are "Youtu-GraphRAG: Vertically Unified Agents for Graph Retrieval-Augmented Complex Reasoning", which aims to evaluate whether the retrieval-augmented generation (RAG) system relies on retrieval to obtain evidence when entities are anonymized.

This dataset is drawn from four classic novels: Water Margin, Dream of the Red Chamber, Moby-Dick, and Middlemarch. It covers both question-answer pairs and text snippets, and is available in both Chinese and English. The question-answer portion contains approximately 1,397 questions, including general questions and answers, multiple-choice questions, and entity anonymization recovery tasks, with questions categorized as simple and complex. The text portion provides text paragraphs as retrieval corpus to support question-answering tasks. It is suitable for RAG model evaluation, complex multi-hop reasoning research, knowledge question-answering system development, and entity anonymization and recovery tasks.