Collatz: notebook records
Working notebook synthesis, not a publication-reviewed article. Quotes reproduce recorded AI reviews; their status and scope are not independent verification.
Stopping times, records and random-walk models
This page is a navigation guide to the existing public records. A dedicated synthesis has not yet been curated. The notebook’s article list remains separate.
Why the evidence matters
These directed relationships restate scoped AI review claims. Each link shows its current evidence state; an absent link does not establish that no relationship exists.
- supports: Recomputing the listed trajectories addresses finite-data objections and transcription errors. It leaves the completeness of the list as a separate obligation.
- depends on: A correct calculation on every listed row can still miss an unlisted record. This dependency explains why finite validation does not establish global coverage.
- contradicts: The reported failed rows block this sufficient certificate. They do not refute equality of the record lists itself.
- contradicts: Two starts with the same four-bit prefix and different first-hit times expose information lost by the prefix. A first-passage estimate needs an additional argument.
- generalizes: The draft explicitly identifies d=1 as the 3x+1 map. This connects the definitions and explains why methods may be compared; results for other d do not automatically prove a Collatz claim.