FrontierCS
View and compare model performance on the FrontierCS benchmark.
Algorithmic Problems, covering Optimization tasks, Constructive tasks, and Interactive tasks, adapted from programming contests but rewritten into open-ended, partially scored variants.
Research Problems, spanning six major CS domains: OS (Operating Systems), HPC (High-Performance Computing), AI (Artificial Intelligence research tasks), DB (Databases), PL (Programming Languages), and Security (cybersecurity and vulnerability analysis).