BENCHPRO, has sparked heated discussions regarding the programming capabilities of artificial intelligence. In this test, the solution rates for GPT-5, Claude Opus 4.1, and Gemini 2.5 were 23.3%, 22.7 ...
Arai, H. , Yamamoto, H. , Akiba, Y. , Takayama, J. , Aiba, S. and Arai, R. (2025) A Voxel-Based Morphometric Pilot Study of ...