利用局部内蕴维度表征大型语言模型生成内容的真实性
Characterizing Truthfulness in Large Language Model Generations with Local Intrinsic Dimension
AI总结：
本文提出利用LLM内部激活的局部内蕴维度（LID）量化并预测生成文本真实性，在四个QA数据集上验证其有效性，并揭示内蕴维度与模型层、自回归建模及训练的关系。
摘要：
我们研究如何表征和预测大型语言模型（LLM）所生成文本的真实性，这是在人类与LLM之间建立信任的关键一步。尽管已有若干基于熵或语言化不确定性的方法被提出用于校准模型预测，但这些方法在应用于LLM的生成任务时往往难以处理、对超参数敏感且可靠性较低。本文建议考察模型内部激活，并利用模型激活的局部内蕴维度（LID）来量化LLM的真实性。通过在四个问答（QA）数据集上的实验，我们证明了所提方法的有效性。此外，我们研究了LLM中的内蕴维度及其与模型层、自回归语言建模和LLM训练之间的关系，揭示内蕴维度可以成为理解LLM的一种强有力方法。
英文摘要：
We study how to characterize and predict the truthfulness of texts generated from large language models (LLMs), which serves as a crucial step in building trust between humans and LLMs. Although several approaches based on entropy or verbalized uncertainty have been proposed to calibrate model predictions, these methods are often intractable, sensitive to hyperparameters, and less reliable when applied in generative tasks with LLMs. In this paper, we suggest investigating internal activations and quantifying LLM's truthfulness using the local intrinsic dimension (LID) of model activations. Through experiments on four question answering (QA) datasets, we demonstrate the effectiveness ohttps://info.arxiv.org/help/prep#abstractsf our proposed method. Additionally, we study intrinsic dimensions in LLMs and their relations with model layers, autoregressive language modeling, and the training of LLMs, revealing that intrinsic dimensions can be a powerful approach to understanding LLMs.
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。
