重新思考LLM幻觉检测的评估：一种需求清单，一个新的基于RAG的基准，新的见解
Rethinking Evaluation for LLM Hallucination Detection: A Desiderata, A New RAG-based Benchmark, New Insights
AI总结：
本文提出了一种基于RAG的新型基准T RIVIA+，填补了现有基准在长上下文和噪声标签方面的空白，并通过实验揭示了当前检测器在RAG基准上的改进空间。
摘要：
幻觉，广义上指由LLM生成的不忠实、伪造或不一致的内容，具有广泛影响。因此，大量工作致力于检测LLM幻觉以及设计评估这些检测器的基准数据集。本文首先建立了幻觉检测基准（HDBs）的属性需求清单，通过这一清单审视现有HDBs，发现它们均未满足所有属性。我们识别出两个主要缺口：（1）基于RAG的有根基准在长上下文方面严重不足（部分原因在于长度阻碍人工标注）；（2）现有基准未提供现实的标签噪声以测试检测器，尽管现实使用场景常面临标签噪声问题。为填补这些缺口，我们构建并开源了一个新的基于RAG的HDB，称为T RIVIA+，其经过严格的人工标注过程。值得注意的是，我们的基准展示了所有理想的属性，包括（1）T RIVIA+包含文献中最长的上下文样本；（2）我们设计并分享了四组具有不同样本依赖性和样本独立性噪声方案的噪声标签。最后，我们在RAG基准上，包括我们的T RIVIA+，使用流行的SOTA检测器进行了实验，揭示了新的见解：（i）当前检测器在RAG基准上仍有改进空间；（ii）基本的LLM-as-a-Judge基线表现良好；（iii）标签噪声会阻碍检测性能。我们期望我们的发现以及我们提出的基准将激励和促进对基于RAG任务的幻觉检测研究。
英文摘要：
Hallucination, broadly referring to unfaithful, fabricated, or inconsistent content generated by LLMs, has wide-ranging implications. Therefore, a large body of effort has been devoted to detecting LLM hallucinations, as well as designing benchmark datasets for evaluating these detectors. In this work, we first establish a desiderata of properties for hallucination detection benchmarks (HDBs) to exhibit for effective evaluation. A critical look at existing HDBs through the lens of our desiderata reveals that none of them exhibits all the properties. We identify two largest gaps: (1) RAG-based grounded benchmarks with long context are severely lacking (partly because length impedes human annotation); and (2) Existing benchmarks do not make available realistic label noise for stress-testing detectors although real-world use-cases often grapple with label noise due to human or automated/weak annotation. To close these gaps, we build and open-source a new RAG-based HDB called T RIVIA+ that underwent a rigorous human annotation process. Notably, our benchmark exhibits all desirable properties including (1) T RIVIA+ contains samples with the longest context in the literature; and (2) we design and share four sets of noisy labels with different, both sample-dependent and sampleindependent, noise schemes. Finally, we perform experiments on RAG-based HDBs, including our T RIVIA+, using popular SOTA detectors that reveal new insights: (i) ample room remains for current detectors to reach the performance ceiling on RAG-based HDBs, (ii) the basic LLM-as-a-Judge baseline performs competitively, and (iii) label noise hinders detection performance. We expect that our findings, along with our proposed benchmark 1 , will motivate and foster needed research on hallucination detection for RAG-based tasks.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
