大型语言模型中幻觉的综合分类法
A comprehensive taxonomy of hallucinations in Large Language Models
AI总结：
本报告提出LLM幻觉的综合分类法，涵盖定义、内在/外在类型、表现、成因、评估与缓解策略，强调幻觉理论必然性，主张未来聚焦检测、缓解与人类监督。
摘要：
大型语言模型（LLMs）彻底改变了自然语言处理，然而它们产生幻觉的倾向——生成看似合理但事实错误或捏造的内容——仍然是一个关键挑战。本报告提供了LLM幻觉的综合分类法，首先给出正式定义和理论框架，该框架认为幻觉在可计算的LLM中具有内在必然性，无论架构或训练如何。报告探讨了核心区分，区分了内在幻觉（与输入上下文矛盾）和外在幻觉（与训练数据或现实不一致），以及事实性（绝对正确性）和忠实性（对输入的遵循）。报告随后详细说明了具体表现，包括事实错误、上下文和逻辑不一致、时间迷失、伦理违规，以及跨代码生成和多模态应用等领域中的任务特定幻觉。报告分析了根本原因，将其分类为数据相关问题、模型相关因素和提示相关影响。此外，报告考察了影响幻觉感知的认知和人为因素，调查了用于检测的评估基准和指标，并概述了架构和系统性缓解策略。最后，报告介绍了用于监控LLM发布和性能的基于网络的资源。本报告强调了LLM幻觉的复杂、多方面性质，并强调鉴于其理论上的必然性，未来的努力必须集中于稳健的检测、缓解和持续的人类监督，以在关键应用中实现负责任和可靠的部署。
英文摘要：
Large language models (LLMs) have revolutionized natural language processing, yet their propensity for hallucination, generating plausible but factually incorrect or fabricated content, remains a critical challenge. This report provides a comprehensive taxonomy of LLM hallucinations, beginning with a formal definition and a theoretical framework that posits its inherent inevitability in computable LLMs, irrespective of architecture or training. It explores core distinctions, differentiating between intrinsic (contradicting input context) and extrinsic (inconsistent with training data or reality), as well as factuality (absolute correctness) and faithfulness (adherence to input). The report then details specific manifestations, including factual errors, contextual and logical inconsistencies, temporal disorientation, ethical violations, and task-specific hallucinations across domains like code generation and multimodal applications. It analyzes the underlying causes, categorizing them into data-related issues, model-related factors, and prompt-related influences. Furthermore, the report examines cognitive and human factors influencing hallucination perception, surveys evaluation benchmarks and metrics for detection, and outlines architectural and systemic mitigation strategies. Finally, it introduces web-based resources for monitoring LLM releases and performance. This report underscores the complex, multifaceted nature of LLM hallucinations and emphasizes that, given their theoretical inevitability, future efforts must focus on robust detection, mitigation, and continuous human oversight for responsible and reliable deployment in critical applications.
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。
