用检索+推理检测大模型幻觉，适合生产环境长文本应用。
ORION Grounded in Context: Retrieval-Based Method for Hallucination Detection
- 结合检索与自然语言推理，通过小模型判断生成内容是否符合上下文。
- 在RAGTruth任务中达到0.83的F1分数，性能媲美大规模训练模型。
- 轻量级设计适合摘要、抽取等多样场景，可直接部署于生产系统。
尽管在基于上下文的内容生成方面取得进展，基于大语言模型的应用仍存在幻觉问题。我们提出“Grounded in Context”——Deepchecks ORION家族中的轻量级评估框架，用于检测生成内容中的事实性错误。该方法受RAG架构启发，融合检索与自然语言推断（NLI）模型，采用仅支持512词元上下文窗口的编码器模型，判断前提与假设间的事实一致性。在RAGTruth的数据集响应级别分类任务中，该框架实现0.83的F1分数，性能与针对该数据集训练的模型相当，且优于所有使用相似规模模型的可比框架。
原文摘要 · Abstract (English)
Despite advancements in grounded content generation, production Large Language Models (LLMs) based applications still suffer from hallucinated answers. We present "Grounded in Context" - a member of Deepchecks' ORION (Output Reasoning-based InspectiON) family of lightweight evaluation models. It is our framework for hallucination detection, designed for production-scale long-context data and tailored to diverse use cases, including summarization, data extraction, and RAG. Inspired by RAG architecture, our method integrates retrieval and Natural Language Inference (NLI) models to predict factual consistency between premises and hypotheses using an encoder-based model with only a 512-token context window. Our framework identifies unsupported claims with an F1 score of 0.83 in RAGTruth's response-level classification task, matching methods that trained on the dataset, and outperforming all comparable frameworks using similar-sized models.
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。