MechReason：机械工程多图像多跳推理基准
原标题：MechReason: Benchmarking Multi-Image Multi-Hop Reasoning in Mechanical Engineering
AI 摘要
研究者提出 MechReason，一个面向机械工程的多图像多跳推理基准，源自真实机械工程论文，包含 12k 问答对、21k 视觉材料和显式推理链标注，覆盖解释、预测、设计、诊断四个维度共八类任务。该基准采用四阶段构建流程，通过遮蔽后验验证信息来防止模型走捷径，并经过多模态质量验证。实验显示最先进的多模态大模型仅达到 62.89% 准确率，错误主要来自推理链缺陷、证据定位失败和有损答案压缩，而非简单幻觉。
正文节选
MechReason: Benchmarking Multi-Image Multi-Hop Reasoning in Mechanical Engineering Abstract Despite significant progress in general visual question answering and cross-modal understanding, multimodal large language models still face a pronounced gap in evaluation for complex reasoning within the mechanical engineering domain. Existing benchmarks predominantly focus on rudimentary tasks such as drawing recognition, CAD interpretation, or single-chart querying, falling short of assessing whether m