A025529: a prime-cube counterexample
Abstract
The prime cube 16843^3 refutes the composite-solution classification for A025529.
Definition 1.1 (The LCM-weighted harmonic sum).
Formalization. D5/S3/ArithSums/A025529PrimeCubeRefutation.A (✓ std3).
Citation. Amiram Eldar; Thomas Ordowski; OEIS Foundation Inc. (2019). OEIS A025529: the LCM-weighted harmonic sum and composite-solution conjecture. URL: https://oeis.org/A025529.
Commentary.
Let L(m) be the least common multiple of 1 through m. Define A(m) as the sum of the exact natural quotients L(m)/k for 1<=k<=m. Under the rational embedding this is L(m) times harmonic(m), as proved inside the lifting argument. The conventions are L(0)=1, harmonic(0)=0 and A(0)=0.
Definition 1.2 (The composite-solution predicate).
Formalization. D5/S3/ArithSums/A025529PrimeCubeRefutation.PrimeSquareOnly (✓ std3).
Citation. Amiram Eldar; Thomas Ordowski; OEIS Foundation Inc. (2019). OEIS A025529: the LCM-weighted harmonic sum and composite-solution conjecture. URL: https://oeis.org/A025529.
Commentary.
PrimeSquareOnly is the formal encoding of the composite-solution conjecture in the 2019 comment on OEIS A025529: every composite n with n dividing A(n-1) is the square of a prime greater than three. The conjunction 1<n and not Prime(n) expresses compositeness.
Theorem 1.3 (Lifting a short harmonic congruence).
Proof. Machine-checked in Lean as D5/S3/ArithSums/A025529PrimeCubeRefutation.prime_cube_divides (✓ std3). ∎
Source. Repository-derived.
Acknowledgement. Amiram Eldar; Thomas Ordowski; OEIS Foundation Inc. (2019). OEIS A025529: the LCM-weighted harmonic sum and composite-solution conjecture. URL: https://oeis.org/A025529.
Commentary.
For a prime p>3, work in the subring of rationals whose reduced denominators are not divisible by p. Write U(N) for the reciprocal sum over 1<=k<N with p not dividing k. Reflection k maps to N-k proves U(N)/N belongs to this subring whenever p divides N. Partitioning denominators gives p^2 H(p^3-1)=H(p-1)+p U(p^2)+p^2 U(p^3). The short ring congruence implies H(p-1)/p^3 belongs to the subring, by clearing the unit denominator (p-1)! abstractly. Since p^2 divides L(p^3-1), the rational identity for A transfers this to the asserted integer divisibility. The ring ZMod(p^3) is not treated as a field: every required denominator is proved to be a unit.
Theorem 1.4 (The complete classification is false).
Proof. Machine-checked in Lean as D5/S3/ArithSums/A025529PrimeCubeRefutation.prime_square_only_refuted (✓ std3). ∎
Source. Repository-derived.
Acknowledgement. Amiram Eldar; Thomas Ordowski; OEIS Foundation Inc. (2019). OEIS A025529: the LCM-weighted harmonic sum and composite-solution conjecture. URL: https://oeis.org/A025529.
Commentary.
The proof establishes that 16843 is prime and checks all 16842 reciprocal terms modulo 16843^3, obtaining zero. The lifting theorem then proves 4778134229107 divides A(4778134229106). This integer equals 16843^3, is composite, and cannot equal the square of any prime. The finite computation occurs in the proof of the negation. Neither the trillion-term harmonic sum nor the large LCM is evaluated. No exact valuation-three assertion or all-exponent valuation identity is used.
References
- Truth anchor:
D5/S3/ArithSums/A025529PrimeCubeRefutation.A - Truth anchor:
D5/S3/ArithSums/A025529PrimeCubeRefutation.PrimeSquareOnly - Truth anchor:
D5/S3/ArithSums/A025529PrimeCubeRefutation.prime_cube_divides - Truth anchor:
D5/S3/ArithSums/A025529PrimeCubeRefutation.prime_square_only_refuted