Keyboard shortcuts

Press ← or → to navigate between chapters

Press ? to show this help

Press Esc to hide this help


slug: oeis-a133901-palindrome-square-reversal-infinitude bibkey: brockhaus2007a133901 doi: null url: https://oeis.org/A133901 triage: theorem motivation_gids:

  • D5/S1/Digit/Admissibility/PalindromeSquareReversalInfinitude

Palindrome-square reversal infinitude in A133901

Problem

OEIS A133901, NAME (verbatim):

Numbers in A128921 whose square is not a palindrome.

OEIS A128921, NAME (verbatim):

Palindromes m such that reverse of m^2 is also a square.

OEIS A133901, COMMENT (verbatim):

Conjecture: Sequence is infinite.

Thus membership means a decimal palindrome p whose square is not a palindrome and whose reversed square is a perfect square. In the formal module, rev10 is the numeric value of the reversed decimal string; it is distinct from the digit-list reversal D in D5/S3/Arith/SumInConcatenation.lean.

Motivation

Brockhaus and Seidov recorded this unboundedness conjecture in 2007. Proving the full statement produces an explicit member above every natural bound and resolves the conjectured infinitude, rather than extending the finite table.

Gap

The surfaces recorded in preregistration issue #7465 and checked by the probe on 2026-09-13 were all three OEIS revisions, exact identifier searches on arXiv, OpenAlex, MathOverflow, and GitHub, and the pinned Lean library search surfaces. Revision #1 introduces the conjecture; revisions #2 and #3 only change author formatting. The four exact identifier searches each returned zero relevant proof or refutation hits. Fried’s 2025 and 2026 JIS sequence lists do not contain A133901 according to the GPT Pro search seat; that reading is ASSUMED-UNVERIFIED at the level of the two papers’ full text. This is a bounded search report, not an exhaustive literature or priority claim.

Route

Given a bound B, put j = B + 5, t = 10^j, and

P_j = 100t^4 + 110t^3 + 90t^2 + 11t + 1,

Q_j = 100t^4 + 110t^3 - 9t^2 + 11t + 1.

The decimal representation of P_j is the (4j+3)-digit palindrome

1 0^(j-1) 11 0^(j-1) 9 0^(j-1) 11 0^(j-1) 1.

The private block-digit reconstruction lemma digits_of_pow_blocks gives the decimal strings of P_j, P_j^2, and Q_j^2: P_j uses five blocks, while each square uses nine five-digit blocks separated by j-5 zeros. Decimal reversal is therefore block-order reversal together with in-block reversal. The resulting identity is rev10(P_j^2) = Q_j^2. Since P_j - Q_j = 99t^2 > 0, it follows that rev10(P_j^2) = Q_j^2 < P_j^2, so P_j^2 is not a palindrome. Finally, P_j > B, and the construction supplies a member above every bound.

Falsifier

A natural bound B above which no A133901 member exists would contradict the theorem. Equivalently, failure of palindrome membership, the reversed square identity, or the strict square inequality for any constructed P_j would invalidate the route.

Evidence

  • Lean module: D5/S1/Digit/Admissibility/PalindromeSquareReversalInfinitude.lean.
  • Main theorem: brockhaus_seidov_a133901.
  • The theorem’s axiom report is std3: propext, Classical.choice, and Quot.sound.
  • The orchestrator checked the family for j = 5..12 and exhaustively searched p < 10^6, finding exactly {33, 99}.
  • The probe independently repeated the same family range and exhaustive search, with the same results.

Triage

theorem. The Lean theorem proves an A133901 member exists above every natural bound, establishing the full conjectured infinitude.

ASSUMED-UNVERIFIED

The absence of A133901 from the full texts of Fried’s 2025 and 2026 JIS papers was supplied by the GPT Pro search seat and was not independently rechecked by this implementation seat. The literature search is bounded to the dated surfaces in Gap. No exhaustive literature or first-publication claim follows. Source-to-Lean identification is not itself kernel-checked.