claimroutineinferred_dep
Fresh isomorphisms make genericity typical
Fresh isomorphisms make the next domain nongeneric with probability ≤
feeds the per-level rank-drop bound
Statement (paper)
For every , conditional on and , the next domain is a uniformly random subspace of of the fixed rank , so that, with the indicator that and is nongeneric, .
In plain English
Conditional on the history and the chosen child , the next domain is a uniformly random subspace of of fixed rank . Hence the chance that is nongeneric with positive rank is at most , i.e. .
Why it's here
The subspace is determined by the history and the child, while the child's isomorphism is fresh randomness, so applying randomizes the subspace uniformly. The random-subspace genericity lemma then bounds the probability of nongeneric by , and this feeds the per-level expected rank-drop bound that uses it.
Classified routine:
Follows directly from the fact that a uniformly random linear isomorphism maps a fixed subspace to a uniformly random subspace of the same rank, combined with the earlier random-subspace genericity lemma (lem-latent-generic).
Uses
Used by
Sources
PaperOpen
- Section
- §3.4 Linear lifts and loss of rank
- Page
- —
- Label
—- Line
sections/02-latent.tex:547
Leannot aligned
Phase 2 aligns this node to a Lean declaration.
Friction
- infoinferred_depoutline: "Fresh isomorphisms make genericity typical" rests on lem:latent-generic
- infoinferred_depoutline: "Fresh isomorphisms make genericity typical" rests on lem:latent-lift-factorization