Can frontier AI agents achieve Recursive Self-Improvement by turning a weak method into one that performs better on hidden data?
Introducing RSI-Exam ๐งต: 88 executable research tasks across 6 domains, now open for task contributions from every field๐งโ๐ฌ.
Its task bank spans virtual cells, TPU kernels, chip design, quantitative finance, agent harnesses, model distillation, and much more.
Across these tasks, every rollout follows the same core protocol:
๐ An agent inherits a working method and iterates using visible data.
๐ Only its final artifact enters a fresh-container evaluation on the hidden set.
๐ The result so far: Opus 5 leads the 88-task leaderboard with a mean hidden-set score of 0.464.
๐ข Want to help build the next release? Author an executable task, review one, or audit a rollout. Substantial contributions qualify for paper authorship.
Special thanks to
@XinyeYee for the tremendous effort in leading this project.
Explore:
Blog:
Contribute:
GitHub: