A novel research infrastructure, Intern-Atlas, has been introduced to facilitate the representation of methodological evolution in AI science. This framework addresses the limitations of existing document-centric research infrastructure by providing explicit representations of how research methods emerge, adapt, and build upon one another. Intern-Atlas captures the structured relationships between methods, enabling a more comprehensive understanding of the evolution of AI research. By doing so, it has the potential to support the development of more advanced AI-driven research agents1. The implications of this development extend beyond the AI research community, as it may influence the trajectory of AI-driven scientific discoveries. This matters to practitioners because it can enhance the transparency and reproducibility of AI research, ultimately contributing to more robust and reliable AI systems.