DHCP：通过大视觉语言模型中的跨模态注意力模式检测幻觉
DHCP: Detecting Hallucinations by Cross-modal Attention Pattern in Large Vision-Language Models
AI总结：
针对大视觉语言模型存在的对象、属性、关系类幻觉问题，提出基于跨模态注意力模式差异的轻量级幻觉检测方法DHCP，无需额外训练或推理步骤即可实现显著检测效果，为提升模型可靠性提供新视角。
摘要：
大视觉语言模型（LVLMs）已在复杂多模态任务上展现出卓越性能，但仍存在严重的幻觉问题，包括对象幻觉、属性幻觉和关系幻觉。为准确检测这些幻觉，我们研究了幻觉状态与非幻觉状态下跨模态注意力模式的差异。利用这些差异，我们开发了一种可识别幻觉的轻量级检测器。我们提出的方法——DHCP（Detecting Hallucinations by Cross-modal Attention Patterns，基于跨模态注意力模式的幻觉检测），实现简单，无需额外的LVLM训练或额外的LVLM推理步骤。实验结果表明，DHCP在幻觉检测上取得了显著性能。DHCP为LVLM中幻觉的识别与分析提供了新见解，有助于提升这类模型的可靠性与可信度。代码可在https://github.com/btzyd/DHCP获取。
英文摘要：
Large vision-language models (LVLMs) have demonstrated exceptional performance on complex multimodal tasks. However, they continue to suffer from significant hallucination issues, including object, attribute, and relational hallucinations. To accurately detect these hallucinations, we investigated the variations in cross-modal attention patterns between hallucination and non-hallucination states. Leveraging these distinctions, we developed a lightweight detector capable of identifying hallucinations. Our proposed method, Detecting Hallucinations by Cross-modal Attention Patterns (DHCP), is straightforward and does not require additional LVLM training or extra LVLM inference steps. Experimental results show that DHCP achieves remarkable performance in hallucination detection. By offering novel insights into the identification and analysis of hallucinations in LVLMs, DHCP contributes to advancing the reliability and trustworthiness of these models. The code is available at https://github.com/btzyd/DHCP.
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。
