🚀NEW COURSEVibe Coding AI Apps with Claude Code 🤖✨Enroll now
Evaluation

Meta^n

First page
Meta^n
Paper summary

Systems that edit themselves have to leave part of their own editing machinery untouched to stay stable, which caps realized meta-depth at roughly two. Meta^n keeps the meta-operation fixed and recurses on its input instead, applying one operator repeatedly to its own products and letting convergence set the depth rather than fixing it in advance. Across two backbones it outperforms prior self-improving agents on all eight benchmark families, and on ARC-AGI-2 it is the only method scoring above zero.

Ask this paper

Every Monday
Get next week’s papers.
Subscribe on Substack