RESEARCH / Focused target
Weak Ascent Sequences Avoiding 210 and the Semi-Baxter Numbers
RELEASED FOUNDATIONS / PROPOSED CONNECTION
Research connections
All source anchors (1)
- D5/S3/Combinatorics/WeakAscent/WeakAscentSemiBaxter.resultAbsent from this graph
Problem
The question
Beáta Bényi, Toufik Mansour and José L. Ramírez, Pattern Avoidance in Weak Ascent Sequences, arXiv:2309.06518v4 (Discrete Math. Theor. Comput. Sci. 26:1, 2024), Section 3, Conjecture 3.1:
The sequence w_210(n) coincides with the sequence A117106.
The paper adds that A117106 enumerates permutations avoiding the vincular pattern 2-41-3. Here w_210(n) counts the weak ascent sequences of length n with no indices i < j < k and e_i > e_j > e_k.
Motivation
Our foothold
The theorem D5/S3/Combinatorics/WeakAscent/WeakAscentSemiBaxter.result establishes, for every n, that
the 210-avoiding weak ascent sequences of length n and the permutations of [n] avoiding 2-41-3 are
equinumerous; both are counted by 1, 1, 2, 6, 23, 104, 530, 2958, 17734, … for n = 0, 1, ….
Gap
Missing bridges
Pre-registration issue 11728 records the literature screen: none of the located papers citing arXiv:2309.06518 treats this conjecture, and OEIS A117106 records no proof. This is a bounded negative finding.
Route
Proposed approach
- For a nonempty 210-avoiding weak ascent sequence put H = 1 + wasc, M = max and D = the largest entry that is the smaller term of some inversion (0 if none). Appending x keeps the class exactly when D ≤ x ≤ H; moreover H ≥ M + 1 and the last entry is D or M.
- Labelling a sequence by (M − D + 1, H − M) when it ends at its maximum and by (M − D, H − M + 1) otherwise, the labels of the children are those of the semi-Baxter succession rule (h, k) → (1, k + 1), …, (h, k + 1), (h + k, 1), …, (h + 1, k) with root (1, 1).
- A 2-41-3-avoiding permutation stays in the class when its maximum is deleted; inserting a new maximum is allowed exactly at the active sites, and labelling by the active sites before and after the maximum gives the same succession rule.
- Induction on the number of steps shows that every object with label (h, k) has the same number of descendants at each depth in both trees, so both classes have T(n − 1, 1, 1) members of length n ≥ 1.
Falsifier
What would falsify this route
The statement would fail if some admissible append or active insertion produced a child label outside the semi-Baxter rule, or if two distinct parents produced the same child.
Evidence
Evidence to collect
The append criterion, the state updates and the children labels were checked on every 210-avoiding weak ascent sequence and every 2-41-3-avoiding permutation of length at most 11; the counts agree through n = 11.
Triage
Scope assessment
theorem; the statement is Conjecture 3.1 of arXiv:2309.06518 and is quantified over every n.
ASSUMED-UNVERIFIED
Unverified assumptions
The literature screen is limited to the located citing papers, the OEIS entry, arXiv and GitHub searches and the repository checks recorded above.