The moment of capture in a quadratic recurrence modulo 3
Fix k ≥ 2 and let x1 = … = xk-1 = 0, xk = 1 and xn = xn-1 + xn-k2 in Z/3. For many k the sequence wanders for a very long time and then settles into a cycle of length 4. Finch conjectured (arXiv:2610.34300, Conjecture 23) that the settling always happens in the same way: if z is the end of the first run of zeros spaced exactly four apart that is long enough to force the cycle, and μ is the pre-period, then z − μ = 6. We prove this for every k ≥ 3 and show that it fails for k = 2. The argument is formalised in Lean 4 with Mathlib. The archive holds the LaTeX source, verification scripts, mutation suites, an independent check and the pinned Lean project (code under the MIT licence).
Authors
- Vamshi Jandhyala (ORCID: https://orcid.org/0009-0006-6708-5805)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-18
- DOI
- https://doi.org/10.5281/zenodo.24352080
- Primary Topic
- Mathematical Dynamics and Fractals
- Type
- preprint