When Models Manipulate Manifolds: The Geometry of a Counting Task

Wes Gurnee, Emmanuel Ameisen, Isaac Kauvar, Julius Tarng, Adam Pearce, Chris Olah, Joshua Batson
Paper From BibTeX import
Transformer Circuits Thread, 2025

Notes

Their finding that next-token training induces low-dimensional internal geometry for structural variables gives riva2026task an empirical counterpart to its topological convergence prediction.

References

No references yet.