The ω₂ exponent: specification and Appendix-B cross-check #
GQ2.omega2Exp n is our concrete integer representative of the profinite idempotent ω₂ modulo
n. This file proves it satisfies the two congruences that define ω₂ — ≡ 1 on the 2-part and
≡ 0 on the odd part — and cross-checks it against the paper's Appendix B serialization.
The odd part of n divides omega2Exp n (the "ω₂ ≡ 0 on the odd part" condition).
omega2Exp n ≡ 1 modulo the 2-part 2 ^ v₂(n) (the "ω₂ ≡ 1 on the 2-part" condition),
for n with a nontrivial 2-part. Uses Euler's theorem: the odd part is a unit mod 2^a, and
2^(a-1) = φ(2^a).
Compatibility of the ω₂ exponents across levels. For N ∣ M (M ≠ 0), the exponents
at levels M and N agree modulo N: omega2Exp M ≡ omega2Exp N [MOD N]. This is the
coherence making the family (omega2Exp N)_N a well-defined element ω₂ of ℤ̂ = lim ℤ/N
(see GQ2.omega2): both sides are ≡ 1 on the 2-part of N and ≡ 0 on its odd part, and CRT
combines the two congruences over the coprime factorisation N = 2^{v₂ N} · (N / 2^{v₂ N}).
ω₂ is well-defined via any exponent multiple. For x of finite order dividing N
(N ≠ 0), x ^ (omega2Exp N) = powOmega2 x. So powOmega2 really is the 2-primary
projection: the choice of modulus orderOf x in its definition is immaterial, as long as the
modulus is a multiple of orderOf x. (This is what makes powOmega2 behave coordinatewise on
products, cf. powOmega2_prod.)
Naturality of ω₂. The 2-primary projection commutes with every group homomorphism (out
of a finite group): f (x ^ ω₂) = (f x) ^ ω₂. This is the structural fact underlying the fact
that the paper's auxiliary words are preserved by quotient maps (needed for Lemma 2.1).
Appendix B, exact match. Our computable representative omega2Exp, evaluated at the
paper's modulus M = 85667662080 = 2⁸·3²·5·7·11·13·17·19·23, reproduces the Appendix-B
serialization 40491355905 exactly (not merely up to the defining congruences). This certifies
that the definition omega2Exp, and not just the hard-coded residue of omega2_appendixB, agrees
with the paper. Proved with only the standard axioms: the 2-adic valuation v₂(M) = 8 is pinned by
p^k ∣ M bounds, and the remaining 334639305 ^ 2⁷ % M is closed by kernel Nat arithmetic.
Paper-tag ledger (auto-generated by paperforge; do not edit) #
- Lemma 2.1 = ⟦lem-subdirect⟧