Sci-Reasoning: A Dataset Decoding AI Innovation Patterns

TL;DR

Sci-Reasoning introduces a novel dataset capturing the reasoning processes behind high-quality AI research papers, enabling analysis of innovation patterns and supporting AI research agent development.

Contribution

The paper presents the first dataset of scientific reasoning links in AI research, identifying key thinking patterns and innovation strategies through a structured, verified pipeline.

Findings

Identified 15 distinct reasoning patterns in AI research.

Three dominant strategies account for over half of the innovation.

Combining multiple reasoning patterns leads to more powerful innovation recipes.

Abstract

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.