超越事实：评估大型语言模型中的意图幻觉
Beyond Facts: Evaluating Intent Hallucination in Large Language Models
AI总结：
本文提出意图幻觉概念与包含20,068题的FAITHQA基准，覆盖仅查询和RAG场景，揭示LLM因遗漏或误解条件而产生意图幻觉，并推出更贴近人工判断的CONSTRAINT SCORE自动评估指标。
摘要：
当面对包含多个条件的复杂查询时，当今的大型语言模型（LLM）往往会生成仅部分满足查询而忽略某些条件的回答。因此，我们提出意图幻觉这一概念。在该现象中，LLM会遗漏（忽略处理某些部分）或误解（对虚构的查询部分作出回应）给定查询中的要素，从而导致意图幻觉生成。为系统评估意图幻觉，我们推出FAITHQA，这是一个用于意图幻觉的新型基准，包含20,068个问题，涵盖仅查询和检索增强生成（RAG）设置，并涉及不同主题与难度。FAITHQA是首个超越事实验证的幻觉基准，专门用于识别意图幻觉的根本原因。通过在FAITHQA上评估多种LLM，我们发现：（1）即使对最先进模型而言，意图幻觉也是一个普遍问题；（2）该现象源于LLM的遗漏或误解。为促进未来研究，我们提出一种自动LLM生成评估指标CONSTRAINT SCORE，用于检测意图幻觉。人工评估结果表明，与基线相比，CONSTRAINT SCORE在意图幻觉方面更接近人类表现。
英文摘要：
When exposed to complex queries containing multiple conditions, today's large language models (LLMs) tend to produce responses that only partially satisfy the query while neglecting certain conditions. We therefore introduce the concept of Intent Hallucination. In this phenomenon, LLMs either omit (neglecting to address certain parts) or misinterpret (responding to invented query parts) elements of the given query, leading to intent hallucinated generation. To systematically evaluate intent hallucination, we introduce FAITHQA, a novel benchmark for intent hallucination that contains 20,068 problems, covering both query-only and retrieval-augmented generation (RAG) setups with varying topics and difficulty. FAITHQA is the first hallucination benchmark that goes beyond factual verification, tailored to identify the fundamental cause of intent hallucination. By evaluating various LLMs on FAITHQA, we find that (1) intent hallucination is a common issue even for state-of-the-art models, and (2) the phenomenon stems from omission or misinterpretation of LLMs. To facilitate future research, we introduce an automatic LLM generation evaluation metric, CONSTRAINT SCORE, for detecting intent hallucination. Human evaluation results demonstrate that CONSTRAINT SCORE is closer to human performance for intent hallucination compared to baselines.
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。
