视觉语言模型能否推断说话者的无知？视觉与语言线索的作用
Can Vision-Language Models Infer Speaker's Ignorance? The Role of Visual and Linguistic Cues
AI总结：
本研究探究视觉语言模型的无知隐含语用推理能力，通过操控视觉与语言线索测试发现，不同模型处理语境线索的方式有差异，Claude整合多线索的能力或体现多模态模型的语用能力涌现。
摘要：
本研究探讨视觉语言模型（VLM，vision-language models）能否执行语用推理，聚焦于无知隐含——即暗示说话者缺乏精确知识的话语。为验证这一点，我们系统操控了两类语境线索：视觉呈现的情境（视觉线索）和基于讨论问题（QUD）的语言提示（语言线索）。仅提供视觉线索时，GPT-4o、Gemini 1.5 Pro和Claude 3.5 sonnet这三款前沿VLM的解读很大程度上基于修饰数词的词汇意义。添加语言线索提升语境信息量后，Claude通过整合两类语境线索，展现出更符合人类的推理表现；相比之下，GPT和Gemini更倾向于精确的字面解读。尽管语境线索的影响有所增强，但它们会独立处理每类语境线索，并将其与语义特征对齐，而非开展语境驱动的推理。这些发现表明，尽管各模型处理语境线索的方式存在差异，但Claude整合多类线索的能力，可能标志着多模态模型正在涌现语用能力。
英文摘要：
This study investigates whether vision-language models (VLMs) can perform pragmatic inference, focusing on ignorance implicatures, utterances that imply the speaker's lack of precise knowledge. To test this, we systematically manipulated contextual cues: the visually depicted situation (visual cue) and QUD-based linguistic prompts (linguistic cue). When only visual cues were provided, three state-of-the-art VLMs (GPT-4o, Gemini 1.5 Pro, and Claude 3.5 sonnet) produced interpretations largely based on the lexical meaning of the modified numerals. When linguistic cues were added to enhance contextual informativeness, Claude exhibited more human-like inference by integrating both types of contextual cues. In contrast, GPT and Gemini favored precise, literal interpretations. Although the influence of contextual cues increased, they treated each contextual cue independently and aligned them with semantic features rather than engaging in context-driven reasoning. These findings suggest that although the models differ in how they handle contextual cues, Claude's ability to combine multiple cues may signal emerging pragmatic competence in multimodal models.
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。
