Weakly separated self-affine carpets

The paper states and proves that for diagonally aligned planar self‑affine carpets G satisfying (A1)–(A2)–(B2) and either (W1) or (C1)&(G1)&(W2), one has dim_H(Λ) = lim_{n→∞} HBA(G_n) (Theorem 1.1) with all assumptions clearly spelled out (A1)–(A2)–(B2), (C1),(C2),(C3),(G1), (W1),(W2) . The proof leverages (i) the general upper bound dim_H(Λ) ≤ HBA(G) (and its n-block analogue) due to Pardo‑Simón , (ii) WSC-based uniform overlap controls in 1D projections (Lemmas 2.1, 2.2; and the block-level multiplicity bounds Lemmas 4.6, 4.7) , and (iii) a block concatenation/ε-blocking scheme plus Fekete-type arguments to pass to a limit (see Remark 4.5 and Section 4.3) . By contrast, the candidate solution asserts for every n that HBA(G_n) ≤ dim_H(Λ) via an identification dim_H(ν_p) = D(p,G_n) for Bernoulli measures under WSC; this contradicts the paper’s explicit inequality dim_H(Λ) ≤ HBA(G_n) used as a starting point in the proof of Theorem 1.1 . The candidate also posits a uniform O(1) bound on the number of factorizations in projected semigroups under WSC, but the paper only obtains subexponential/combinatorial multiplicity controls depending on block length (Lemmas 4.6, 4.8) rather than a uniform constant bound . Finally, instead of the asserted almost‑superadditivity with O(1/(n+m)) error, the paper uses a refined blocking plus Fekete’s lemma to produce limits, which is substantively different and avoids the candidate’s unjustified step .