Reefknot：面向多模态大语言模型中关系幻觉评估、分析与缓解的综合基准
Reefknot: A Comprehensive Benchmark for Relation Hallucination Evaluation, Analysis and Mitigation in Multimodal Large Language Models
AI总结：
针对多模态大语言模型中复杂的关系幻觉问题，本文提出综合基准Reefknot进行系统评估，并设计基于置信度的缓解策略，使三个数据集上的幻觉率平均降低9.75%。
摘要：
幻觉问题持续影响多模态大语言模型（MLLMs），现有研究主要针对对象级或属性级幻觉，忽略了需要高级推理的更复杂的关系幻觉。当前的关系幻觉基准缺乏详细评估和有效缓解，且其数据集常因系统化标注过程而存在偏差。为应对这些挑战，我们引入了Reefknot，一个针对关系幻觉的综合基准，包含超过20,000个真实世界样本。我们整合感知与认知视角，系统定义了关系幻觉，并利用Visual Genome场景图数据集构建了基于关系的语料库。对比评估揭示了当前MLLMs在处理关系幻觉方面的显著局限。此外，我们提出一种新颖的基于置信度的缓解策略，在包括Reefknot的三个数据集上将幻觉率平均降低9.75%。我们的工作为实现可信的多模态智能提供了宝贵见解。
英文摘要：
Hallucination issues continue to affect multimodal large language models (MLLMs), with existing research mainly addressing object-level or attribute-level hallucinations, neglecting the more complex relation hallucinations that require advanced reasoning. Current benchmarks for relation hallucinations lack detailed evaluation and effective mitigation, and their datasets often suffer from biases due to systematic annotation processes. To address these challenges, we introduce Reefknot, a comprehensive benchmark targeting relation hallucinations, comprising over 20,000 real-world samples. We provide a systematic definition of relation hallucinations, integrating perceptive and cognitive perspectives, and construct a relation-based corpus using the Visual Genome scene graph dataset. Our comparative evaluation reveals significant limitations in current MLLMs' ability to handle relation hallucinations. Additionally, we propose a novel confidence-based mitigation strategy, which reduces the hallucination rate by an average of 9.75% across three datasets, including Reefknot. Our work offers valuable insights for achieving trustworthy multimodal intelligence.
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。
