按年被提到
20232023 · 1 次2026
大模型与语言模型›幻觉与语言模型局限·arXiv:2312.14183 · cs.CL

On Early Detection of Hallucinations in Factual Question Answering

Ben Snyder, Marius Moisescu, Muhammad Bilal Zafar · 2023
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 2 人首次 2023-12-26 · 最近 2024-01-04
2不同的人
2活跃月份
2讨论串
25被引(OpenAlex)
arXiv 摘要(候选层不写导读)

While large language models (LLMs) have taken great strides towards helping humans with a plethora of tasks, hallucinations remain a major impediment towards gaining user trust. The fluency and coherence of model generations even when hallucinating makes detection a difficult task. In this work, we explore if the artifacts associated with the model generations can provide hints that the generation will contain hallucinations. Specifically, we probe LLMs at 1) the inputs via Integrated Gradients based token attribution, 2) the outputs via the Softmax probabilities, and 3) the internal state via self-attention and fully-connected layer activations for signs of hallucinations on open-ended question answering tasks. Our results show that the distributions of these artifacts tend to differ between hallucinated and non-hallucinated generations. Building on this insight, we train binary classifiers that use these artifacts as input features to classify model generations into hallucinations and non-hallucinations. These hallucination classifiers achieve up to $0.80$ AUROC. We also show that tokens preceding a hallucination can already predict the subsequent hallucination even before it occurs.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

没有评论直接链到它;下面是它被提交的讨论串里回复最多的评论(不计入统计)
讨论串评论kig · 2023-12

“The researchers found that certain artifacts associated with LLM model generations could potentially indicate whether or not a model is hallucinating. Their results showed that the distributions of these artifacts were different between hallucinated and non-hallucinated generations. Using these artifacts, they trained binary classifiers to classify model …”

On Early Detection of Hallucinations in Factual Question Answering原评论 ↗
被提交为 story · 2 次