轻量级开源系统检测大模型幻觉，性能超越GPT-4o
Osiris: A Lightweight Open-Source Hallucination Detection System
- 用含诱导幻觉的多跳问答数据集微调7B模型
- 召回率优于GPT-4o，精度与准确率也具竞争力
- 参数量极小，适合生产环境部署
检索增强生成（RAG）系统因能提升大语言模型（LLM）回答的事实性而被广泛应用。然而，幻觉——即模型输出与给定上下文不符——常导致这些系统无法投入生产。现有检测方法依赖人工评估或闭源模型，均存在成本高、推理慢的问题。本文构建了一个含诱导幻觉的扰动多跳问答数据集，通过在该数据集上监督微调，使7B模型在RAGTruth幻觉检测基准上的召回率超过GPT-4o，同时在精度和准确率上表现良好，且仅使用极少参数。代码已开源。
原文摘要 · Abstract (English)
Retrieval-Augmented Generation (RAG) systems have gained widespread adoption by application builders because they leverage sources of truth to enable Large Language Models (LLMs) to generate more factually sound responses. However, hallucinations, instances of LLM responses that are unfaithful to the provided context, often prevent these systems from being deployed in production environments. Current hallucination detection methods typically involve human evaluation or the use of closed-source models to review RAG system outputs for hallucinations. Both human evaluators and closed-source models suffer from scaling issues due to their high costs and slow inference speeds. In this work, we introduce a perturbed multi-hop QA dataset with induced hallucinations. Via supervised fine-tuning on our dataset, we achieve better recall with a 7B model than GPT-4o on the RAGTruth hallucination detection benchmark and offer competitive performance on precision and accuracy, all while using a fraction of the parameters. Code is released at our repository.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。