构建跨论文科研演进链，自动梳理方法与实验的关联脉络。
ResearchPulse: Building Method-Experiment Chains through Multi-Document Scientific Inference
- 用三个协作智能体拆解任务、建动机-方法图谱、生成实验图表。
- 在语义对齐和结构一致性上优于GPT-4o，7B模型实现高精度。
- 适合追踪领域进展的研究者，尤其关注科研演化路径分析者。
理解科学思想的演变不仅需要总结单篇论文，更需对主题相关文献进行结构化、跨文档推理。本文提出多文档科学推理新任务：从相关论文中提取并对齐动机、方法与实验结果，重构研究发展链条。该任务面临时序对齐松散方法及异构实验表格标准化等挑战。我们提出ResearchPulse框架，基于智能体系统，集成指令规划、科学内容抽取与结构化可视化，包含三个协同智能体：规划智能体负责任务分解，Mmap-Agent构建动机-方法心智图，Lchart-Agent合成实验线图。为支持该任务，我们构建ResearchPulse-Bench，一个带引用感知标注的论文簇基准数据集。实验表明，尽管使用7B规模智能体，系统在语义对齐、结构一致性和视觉保真度上持续优于强基线GPT-4o。数据集已公开于https://huggingface.co/datasets/ResearchPulse/ResearchPulse-Bench。
原文摘要 · Abstract (English)
Understanding how scientific ideas evolve requires more than summarizing individual papers-it demands structured, cross-document reasoning over thematically related research. In this work, we formalize multi-document scientific inference, a new task that extracts and aligns motivation, methodology, and experimental results across related papers to reconstruct research development chains. This task introduces key challenges, including temporally aligning loosely structured methods and standardizing heterogeneous experimental tables. We present ResearchPulse, an agent-based framework that integrates instruction planning, scientific content extraction, and structured visualization. It consists of three coordinated agents: a Plan Agent for task decomposition, a Mmap-Agent that constructs motivation-method mind maps, and a Lchart-Agent that synthesizes experimental line charts. To support this task, we introduce ResearchPulse-Bench, a citation-aware benchmark of annotated paper clusters. Experiments show that our system, despite using 7B-scale agents, consistently outperforms strong baselines like GPT-4o in semantic alignment, structural consistency, and visual fidelity. The dataset are available in https://huggingface.co/datasets/ResearchPulse/ResearchPulse-Bench.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。