通过关联分析量化与归因大型语言模型的幻觉
Quantifying and Attributing the Hallucination of Large Language Models via Association Analysis
AI总结：
本文提出通过关联分析结合幻觉水平量化与原因调查，建立LLMs幻觉率与风险因素的关系，揭示模型在常识记忆、关系推理和指令遵循上的潜在缺陷，为预训练和监督微调提供缓解幻觉的指导。
摘要：
尽管在各种NLP任务上表现出色，大型语言模型（LLMs）仍面临幻觉问题，这威胁了LLMs的可靠性。为衡量LLMs的幻觉水平，以往工作先根据现象相似性对幻觉分类，再量化模型输出包含幻觉内容的比例。然而，这种幻觉率很容易被混杂因素扭曲。此外，该幻觉率无法反映幻觉的原因，因为相似的幻觉现象可能源于不同的来源。为解决这些问题，我们提出通过关联分析结合幻觉水平量化与幻觉原因调查，建立LLMs幻觉率与一组风险因素之间的关系。这样，我们能观察每个风险因素取值下的幻觉水平，检验各风险因素的贡献和统计显著性，同时排除其他因素的混杂效应。此外，通过依据模型能力分类法识别风险因素，我们揭示了常识记忆、关系推理和指令遵循方面的一系列潜在缺陷，这可能进一步为LLMs的预训练和监督微调过程提供指导以缓解幻觉。
英文摘要：
Although demonstrating superb performance on various NLP tasks, large language models (LLMs) still suffer from the hallucination problem, which threatens the reliability of LLMs. To measure the level of hallucination of LLMs, previous works first categorize the hallucination according to the phenomenon similarity, then quantify the proportion that model outputs contain hallucinatory contents. However, such hallucination rates could easily be distorted by confounders. Moreover, such hallucination rates could not reflect the reasons for the hallucination, as similar hallucinatory phenomena may originate from different sources. To address these issues, we propose to combine the hallucination level quantification and hallucination reason investigation through an association analysis, which builds the relationship between the hallucination rate of LLMs with a set of risk factors. In this way, we are able to observe the hallucination level under each value of each risk factor, examining the contribution and statistical significance of each risk factor, meanwhile excluding the confounding effect of other factors. Additionally, by recognizing the risk factors according to a taxonomy of model capability, we reveal a set of potential deficiencies in commonsense memorization, relational reasoning, and instruction following, which may further provide guidance for the pretraining and supervised fine-tuning process of LLMs to mitigate the hallucination.
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。
