InfoCausalQA：模型能否基于信息图执行非显式因果推理？
InfoCausalQA:Can Models Perform Non-explicit Causal Reasoning Based on Infographic?
AI总结：
本研究提出InfoCausalQA基准，通过定量与语义因果推理任务评估视觉语言模型基于信息图的因果推断能力，发现当前模型表现显著低于人类，凸显多模态AI在因果推理上的不足。
摘要：
视觉语言模型（VLMs）的最新进展展示了其在感知和推理方面的出色能力。然而，执行因果推断——人类认知的核心方面——的能力仍未被充分探索，特别是在多模态环境中。在本研究中，我们引入了InfoCausalQA，这是一个旨在评估基于信息图的因果推理的新颖基准，该基准结合了结构化视觉数据与文本语境。该基准包含两项任务：任务1侧重于基于推断数值趋势的定量因果推理，而任务2针对涉及五类因果关系的语义因果推理：原因、结果、干预、反事实和时间。我们手动从四个公共来源收集了494个信息图-文本对，并使用GPT-4o生成了1,482个高质量的多项选择QA对。随后这些问题由人工仔细修订，以确保它们无法仅凭表面线索回答，而是需要真正的视觉基础。我们的实验结果表明，当前的VLMs在计算推理方面表现出有限的能力，而在语义因果推理方面的局限性更为明显。它们的表现显著低于人类，表明在利用基于信息图的信息进行因果推断方面存在巨大差距。通过InfoCausalQA，我们强调了推进多模态AI系统因果推理能力的必要性。
英文摘要：
Recent advances in Vision-Language Models (VLMs) have demonstrated impressive capabilities in perception and reasoning. However, the ability to perform causal inference -- a core aspect of human cognition -- remains underexplored, particularly in multimodal settings. In this study, we introduce InfoCausalQA, a novel benchmark designed to evaluate causal reasoning grounded in infographics that combine structured visual data with textual context. The benchmark comprises two tasks: Task 1 focuses on quantitative causal reasoning based on inferred numerical trends, while Task 2 targets semantic causal reasoning involving five types of causal relations: cause, effect, intervention, counterfactual, and temporal. We manually collected 494 infographic-text pairs from four public sources and used GPT-4o to generate 1,482 high-quality multiple-choice QA pairs. These questions were then carefully revised by humans to ensure they cannot be answered based on surface-level cues alone but instead require genuine visual grounding. Our experimental results reveal that current VLMs exhibit limited capability in computational reasoning and even more pronounced limitations in semantic causal reasoning. Their significantly lower performance compared to humans indicates a substantial gap in leveraging infographic-based information for causal inference. Through InfoCausalQA, we highlight the need for advancing the causal reasoning abilities of multimodal AI systems.
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。
