Three Brillhart–Lehmer–Selfridge primality proofs for Wagstaff numbers

Alexey Dolotov · 2026

Read Three Brillhart–Lehmer–Selfridge primality proofs for Wagstaff numbers by Alexey Dolotov (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-18555.txt.

Introduction

Alexey Dolotov gives, in a 2026 paper, fully checked prime proofs for three Wagstaff numbers, W2617, W10501, and W12391, with 788, 3161, and 3730 digits. The proofs use the Brillhart-Lehmer-Selfridge N-1 test on cyclotomic splits, free of ECPP. Factored content comes from Cunningham project tables, with every factor re-checked, and each cofactor proven outright by APR-CL. A Chua N+1 check also confirms the math at each number.

Catalog record

CREATORAlexey Dolotov
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE29988
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.18555

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Three Brillhart–Lehmer–Selfridge primality proofs for Wagstaff numbers Alexey Dolotov Abstract . The Wagstaff numbers W p = ( 2 p + 1 ) / 3 W_{p}=(2^{p}+1)/3 for odd primes p p are the natural + 1 +1 companions of the Mersenne numbers. Known primality proofs for W p W_{p} with p ≥ 2617 p\geq 2617 (as recorded in the Caldwell/Wagstaff-project archives [ 6 , 5 ] ) rely on the elliptic-curve primality proving algorithm of Atkin–Morain [ 2 ] ; Chebyshev/Lucas-type tests, while available as compositeness criteria, remain conjectural on the sufficiency side. We present fully verified primality proofs of W 2617 W_{2617} (788 digits), W 10501 W_{10501} (3161 digits), and W 12391 W_{12391} (3730 digits), independent of ECPP and relying only on classical N − 1 N-1 machinery. The proofs apply the Brillhart–Lehmer–Selfridge (BLS) N − 1 N-1 criterion [ 3 ] to the cyclotomic decomposition 2 p − 1 − 1 = ∏ d | p − 1 Φ d ​ ( 2 ) 2^{p-1}-1=\prod_{d\mid p-1}\Phi_{d}(2) , harvesting factored content from the Cunningham project tables [ 6 ] (used as evidence) and FactorDB [ 9 ] (used only as a discovery aid, with every retrieved factor re-certified). As an independent check on the ℤ ⁡ [ 2 ] \mathbb{Z}[\sqrt{2}] arithmetic implementation, the Chua N + 1 N+1 congruence ω 3 ( W p + 1 ) / 2 ≡ − 1 ( mod W p ) \omega_{3}^{(W_{p}+1)/2}\equiv-1\pmod{W_{p}} — the a = 3 a=3 case of the Chua framework [ 7 ] with ω 3 = 3 + 2 ​ 2 \omega_{3}=3+2\sqrt{2} , a necessary condition for Wagstaff primality — is verified at each W p W_{p} . BLS N − 1 N-1 requires p − 1 p-1 sufficiently smooth that enough cyclotomic factors Φ d ​ ( 2 ) \Phi_{d}(2) are fully factored. On the factorisation data consulted (Cunningham project tables and FactorDB, January–April 2026), p = 10501 p=10501 and p = 12391 p=12391 are the only exponents above 2617 2617 in the known Wagstaff prime/PRP list meeting this ceiling (Remark 4.4 ). Every cofactor primality is certified unconditionally by APR-CL [ 1 , 8 ] ; the method is independent of the Chebyshev sufficiency conjecture, and every step is reproducible from the archived scripts. 2020 Mathematics Subject Classification Primary 11Y11; Secondary 11A51, 11B39 1. Introduction 1.1. Wagstaff numbers and the ECPP era For an odd prime p , the Wagstaff number is W_{p}\;=\;\frac{2^{p}+1}{3}. The exponents p for which W_{p} is prime form OEIS A000978; the Wagstaff primes themselves, as a sequence of integers, form OEIS A000979. All W_{p} with p\leq 42737 have been proved prime or composite (see Appendix B for the 29 proved exponents); 7 more probable primes are known up to the current computational horizon. In the Caldwell Prime Pages archive [ 5 ] and the Cunningham project tables [ 6 ] as consulted for this paper (January–April 2026), the recorded primality proofs for Wagstaff numbers W_{p} with p\geq 2617 are ECPP proofs in the sense of Atkin–Morain [ 2 ] . We are not aware of previously published BLS N-1 proofs for W_{2617} , W_{10501} , or W_{12391} . The Mersenne counterparts M_{p}=2^{p}-1 admit the Lucas–Lehmer test [ 10 ] , a deterministic standalone primality criterion. No deterministic standalone test of comparable simplicity is known for Wagstaff numbers. A natural Chebyshev/Lucas-type candidate — the single congruence \omega_{3}^{(W_{p}+1)/2}\;\equiv\;-1\pmod{W_{p}},\qquad\omega_{3}=3+2\sqrt{2}, in \mathbb{Z}[\sqrt{2}] (the a=3 case of the Chua framework [ 7 ] ) — is known to hold for every Wagstaff prime, but its sufficiency remains conjectural. In the absence of a deterministic standalone test, the Brillhart–Lehmer–Selfridge (BLS) framework [ 3 ] provides unconditional primality proofs whenever N\pm 1 has a “large enough” completely factored divisor: in the N-1 form used here, one needs a divisor F\mid N-1 with F^{3}>N and with every prime of F independently certified. We present BLS-based primality proofs for the three exponents p\in\{2617,10501,12391\} . 1.2. Motivation The three exponents addressed here are already known prime; the contribution of the present paper is methodological, not existence-of-a-prime. Two points motivate revisiting them via BLS: • Auditability. A BLS certificate is a short JSON record of per-prime witnesses plus a discriminant check; re-verification is a handful of modular exponentiations whose correctness rests on standard Fermat-type arithmetic and on APR-CL for each cofactor. ECPP certificates are authoritative but substantially larger and require elliptic-curve evaluation to replay; a BLS/APR-CL pipeline provides an independent confirmation of the primality claim. • Technique demonstration. The cyclotomic harvesting of factored content from 2^{p-1}-1=\prod_{d\mid p-1}\Phi_{d}(2) for the Wagstaff N-1 side is straightforward in principle but, so far as we are aware, has not been applied in the published literature to the specific exponents p\in\{2617,10501,12391\} . The procedure is generalisable: any Wagstaff exponent p with sufficiently smooth p-1 and sufficient Cunningham coverage is a candidate. 1.3. This paper We give three unconditional BLS primality proofs, one per exponent. The proofs and the cross-check rest on three ingredients: (1) Establishing that Condition (II), \omega_{3}^{(W_{p}+1)/2}\equiv-1\pmod{W_{p}} , holds for every Wagstaff prime W_{p} (Proposition 2.3 ) — this is the a=3 case of the Chua framework [ 7 ] and follows from W_{p}\equiv 3\pmod{8} . (2) For the three target exponents p\in\{2617,10501,12391\} , applying the BLS N-1 criterion (Theorem 3.1 ) to the cyclotomic decomposition 2^{p-1}-1=\prod_{d\mid p-1}\Phi_{d}(2) , harvesting factored portions from the Cunningham project tables [ 6 ] plus direct ECM/ p{-}1 /Pollard- \rho computation. Every cofactor primality is certified by APR-CL [ 1 , 8 ] . (3) Verifying Condition (II) independently at each of the three W_{p} as an auxiliary cross-substrate check on the \mathbb{Z}[\sqrt{2}] arithmetic implementation (a distinct code path from the \mathbb{Z} -only BLS arithmetic). The resulting proofs are short (Section 4 ), fully unconditional, and cleanly separated from the Chebyshev sufficiency conjecture (not used, not needed). All computations are reproducible from the scripts described in Section 5 . 1.4. Notation Throughout, p\geq 5 denotes an odd prime and W_{p}=(2^{p}+1)/3 . For an integer or algebraic integer x coprime to a modulus m , \ord_{m}(x) denotes the multiplicative order of x modulo m ; v_{q}(\cdot) denotes the q -adic valuation. We work in \mathbb{Z}[\sqrt{2}] , with fundamental unit \alpha=1+\sqrt{2} and Chebyshev base \omega_{3}=\alpha^{2}=3+2\sqrt{2} . The Pell sequences (U_{n}) and (V_{n}) are defined by \alpha^{n}=V_{n}/2+U_{n}\sqrt{2} , with the recurrence X_{n+1}=2X_{n}+X_{n-1} and the identity V_{n}^{2}-8U_{n}^{2}=4(-1)^{n} . The Jacobi symbol is written (a/n) . 2. Chebyshev bases and the Chua framework 2.1. The Chua identity For an integer a\neq 0,\pm 1 , set \omega_{a}=a+\sqrt{a^{2}-1} , so that \omega_{a}\bar{\omega}_{a}=1 and \omega_{a}^{n}+\omega_{a}^{-n}=2T_{n}(a) where T_{n} is the n -th Chebyshev polynomial of the first kind — hence the name “Chebyshev base”. Theorem 2.1 (Chua [ 7 ] ; cf. Williams [ 11 , Ch. 4] ) . Let Q be an odd prime with \gcd(a^{2}-1,Q)=1 . Set D=a^{2}-1 , \varepsilon=(D/Q) , \delta=(2(a+1)/Q) . Then \omega_{a}^{(Q-\varepsilon)/2}\;\equiv\;\delta\pmod{Q} in \mathbb{Z}[\sqrt{D}]/(Q) . Proof. Define \gamma=a+1+\sqrt{D} ; then \gamma^{2}=2(a+1)\cdot\omega_{a} in \mathbb{Z}[\sqrt{D}] . Two evaluations of \gamma^{Q} in \mathbb{Z}[\sqrt{D}]/(Q) : • Frobenius: \sqrt{D}^{\,Q}\equiv\varepsilon\sqrt{D} (where \varepsilon=(D/Q)\in\{\pm 1\} ), so \gamma^{Q}\equiv a+1+\varepsilon\sqrt{D} . • Via the square: \gamma^{Q}=\gamma\cdot(\gamma^{2})^{(Q-1)/2}=\gamma\cdot(2(a+1))^{(Q-1)/2}\omega_{a}^{(Q-1)/2}\equiv\gamma\cdot\delta\cdot\omega_{a}^{(Q-1)/2} , where \delta=(2(a+1)/Q) by Euler’s criterion. Equating: \gamma\cdot\delta\cdot\omega_{a}^{(Q-1)/2}\equiv a+1+\varepsilon\sqrt{D} in \mathbb{Z}[\sqrt{D}]/(Q) . Case \varepsilon=+1 . The right-hand side equals \gamma , and \gamma is a unit modulo

Read the complete text