trureturingGitHub MATHEMATICAL ATLAS

RESEARCH / Focused target

Level Sequences Avoiding 101 and 102

Literature status: not recheckedRepository record: proved

RELEASED FOUNDATIONS / PROPOSED CONNECTION

Research connections

0 released anchors
All source anchors (1)

Problem

The question

Toufik Mansour, Wilf Classes for Level Sequences Avoiding Patterns of Length Three, Mathematics 14(11), 1983 (2026), Section 4:

Problem 1. We have that F_{101,102}(x) = (1 − √(1 − 4x))/(2x) − 1.

A level sequence of length n is a sequence w_1 ⋯ w_n of nonnegative integers with w_1 = 0 and w_i ≤ 1 + lev(w_1 ⋯ w_{i−1}), where lev counts adjacent equal entries; the problem asks for the number of level sequences avoiding the patterns 101 and 102.

Motivation

Our foothold

The theorem D5/S3/Combinatorics/LevelSequence/LevelSequenceCatalan.result establishes the statement: for every n ≥ 1 the number of level sequences of length n avoiding 101 and 102 is the Catalan number C_n.

Gap

Missing bridges

Pre-registration issue 12044 records the literature screen: the citation index records no citing work, and the arXiv, OEIS and GitHub searches located no later treatment of the problem. This is a bounded negative finding.

Route

Proposed approach

  1. A word avoids 101 and 102 exactly when every descent is permanent: after w_i > w_j with i < j, every later entry is smaller than w_i.
  2. Words over a k-letter alphabet with this property decompose at the first letter into an upper block and a lower word, which gives a convolution recurrence for their counts f_k.
  3. A level sequence avoiding 101 and 102 splits uniquely into its maximal weakly increasing prefix and a descending sequence of value bands; each band is an arbitrary avoiding word over its interval of values.
  4. The level bound becomes a slack variable on the prefix, and the resulting succession rule is solved by the kernel method; the kernel root satisfies (2 + x)Z^2 − (1 + 2x)Z + x = 0, which yields the Catalan series.

Falsifier

What would falsify this route

The statement would fail if the prefix-and-band decomposition missed or repeated a sequence, if a band could contain a value outside its interval, or if the slack recursion admitted a step that violates the level bound.

Evidence

Evidence to collect

Every lemma, the decomposition in both directions, the slack recursion and the series identities were checked on all level sequences of length at most 11; the counts agree with the Catalan numbers through n = 15.

Triage

Scope assessment

theorem; the statement is the open Problem 1 of the source and is quantified over every n ≥ 1.

The same strong-descent description shows that, over any fixed alphabet [k], the words avoiding 101 and 102 are counted by the recurrence f_k = 1 + x Σ_{a=1}^{k} f_{k−a+1} f_{a−1} with the explicit bivariate generating function obtained in the proof (proved). Whether the slack decomposition gives closed forms for the other pairs and triples of length-three patterns that the source treats only through multiple sums was not examined (open).

ASSUMED-UNVERIFIED

Unverified assumptions

The literature screen is limited to the citation index, arXiv, OEIS and GitHub searches and the repository checks recorded above; the publisher full text was read through ResearchGate, not directly.

Dossier & anchor history

Loading release observations...