SEP 10, 2026 · PREPRINT
The information geometry of large language models is shared, learned, and controllable
arXiv
This is a mechanistic and theoretical study of large language model geometry using mathematical frameworks; it presents exploratory analyses and controlled experiments on model internals without clinical outcomes, hard endpoints, or peer review.
Hypothesis-GeneratingDario Picozzi
Study details
DesignMechanistic study with controlled experimen…
InterventionGeometric control via minimum-disturbance l…
ComparatorEuclidean control; human word choices as re…