arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2601.04577cs.AIcs.LG
收藏

Sci-Reasoning:一个解码AI创新模式的数据集

Sci-Reasoning: A Dataset Decoding AI Innovation Patterns

Orchestra Research(Orchestra研究)

Jiachen Liu, Maestro Harmon, Zechen Zhang

AI总结:

Sci-Reasoning数据集通过分析AI研究中的15种思维模式,揭示了高影响力论文的创新机制,为训练下一代AI研究代理提供结构化推理轨迹。

摘要:

尽管AI创新加速发展,但突破背后的知识过程——研究人员如何识别差距、综合已有工作并生成见解——仍缺乏深入理解。缺乏结构化的科学推理数据阻碍了AI研究代理的系统分析和开发。我们引入了Sci-Reasoning,这是首个捕捉高质量AI研究背后知识综合的数据集。通过社区验证的质量信号和一个由大语言模型加速、人工验证的流程,我们追踪了NeurIPS、ICML和ICLR(2023-2025)中的Oral和Spotlight论文及其关键前驱,以结构化的方式阐述具体的推理链接。我们的分析识别出15种不同的思维模式,其中三种主导策略占52.7%:缺口驱动重构(24.2%)、跨领域综合(18.0%)和表征转移(10.5%)。最强大的创新配方结合了多种模式:缺口驱动重构+表征转移,跨领域综合+表征转移,以及缺口驱动重构+跨领域综合。该数据集使科学进步的定量研究成为可能,并为训练下一代AI研究代理提供了结构化的推理轨迹。

英文摘要:

While AI innovation accelerates rapidly, the intellectual process behind breakthroughs -- how researchers identify gaps, synthesize prior work, and generate insights -- remains poorly understood. The lack of structured data on scientific reasoning hinders systematic analysis and development of AI research agents. We introduce Sci-Reasoning, the first dataset capturing the intellectual synthesis behind high-quality AI research. Using community-validated quality signals and an LLM-accelerated, human-verified pipeline, we trace Oral and Spotlight papers across NeurIPS, ICML, and ICLR (2023-2025) to its key predecessors, articulating specific reasoning links in a structured format. Our analysis identifies 15 distinct thinking patterns, with three dominant strategies accounting for 52.7%: Gap-Driven Reframing (24.2%), Cross-Domain Synthesis (18.0%), and Representation Shift (10.5%). The most powerful innovation recipes combine multiple patterns: Gap-Driven Reframing + Representation Shift, Cross-Domain Synthesis + Representation Shift, and Gap-Driven Reframing + Cross-Domain Synthesis. This dataset enables quantitative studies of scientific progress and provides structured reasoning trajectories for training the next generation AI research agents.

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。

↑