Vroom-Vroom at SHROOM-Visions: 用于检测视觉语言输出中幻觉片段的多评委委员会
Vroom-Vroom at SHROOM-Visions: A Multi-Judge Committee for Detecting Hallucinated Spans in Vision-Language Outputs
浏览论文内容
中文总结 AI 辅助
本文提出多评委委员会方法，通过微调视觉语言模型投票检测幻觉片段，在SHROOM-Visions任务中三种语言排名第一，并发现模型分歧与人类分歧相关。
中文摘要 AI 辅助
本文描述了我们对SHROOM-Visions共享任务的提交，该任务旨在检测和分类四种语言中视觉语言模型输出的幻觉字符片段。我们采用多个微调的视觉语言模型作为独立标注器，并通过字符级多数投票结合它们的片段预测，此外还探索了激活探针。该方法在四种语言中的三种中排名第一，并在每种语言和指标上都登上领奖台。我们的分析表明，不同模型之间的分歧与人类标注者之间的分歧相对应。
英文摘要
This paper describes our submission to the SHROOM-Visions shared task on detecting and classifying hallucinated character spans in vision-language model outputs across four languages. We employ several fine-tuned vision-language models as independent annotators and combine their span predictions through character-level majority voting, and additionally explore activation probes. The approach ranks first in three of four languages and places on the podium in every language and metric. Our analysis indicates that disagreement among diverse models tracks disagreement among human annotators.
发表机构
- VTT Technical Research Centre of Finland（芬兰国家技术研究中心）
机构由 AI 辅助整理，请以论文原文为准。
