Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated


Claim. S. A. Burr, P. Erdős, R. L. Graham and W. Wen-Ching Li, Complete sequences of sets of integer powers, Acta Arith. 77 (1996), no. 2, 133--138 (the source card is burr_1996_complete_sequences_sets_integer_powers). For a set AA of integers greater than 11, Pow(A;s)\mathrm{Pow}(A;s) is the set of powers aja^j with a∈Aa\in A and j≥sj\ge s. Section 3 states the largest integer that is not a sum of distinct elements of Pow(A;1)\mathrm{Pow}(A;1) for four sets AA: 581581 for {3,4,7}\{3,4,7\}, which the authors derive from the Mignotte--Waldschmidt lower bound on ∣3p−4q∣|3^p-4^q|; 111111 for {3,5,7,13}\{3,5,7,13\}; 1616 for {3,6,7,13,21}\{3,6,7,13,21\}; and 7878 for {3,4,5}\{3,4,5\}, the last offered as an example of what the authors call their "limited computational experience" with sets whose reciprocal sum exceeds 11. The paper prints none of the computations. Each of the four sets has ∑a∈A1/(a−1)≥1\sum_{a\in A}1/(a-1)\ge1 (with equality for the first three) and greatest common divisor 11, so each is an admissible tuple of [[problems/diophantine_problems/E0124/_index|Problem 124]], and a sum of distinct elements of Pow(A;1)\mathrm{Pow}(A;1) is a sum ∑iciai\sum_ic_ia_i with ai∈P(di,1)a_i\in P(d_i,1), one term per base.

Covers. The second question at k=1k=1 for the four tuples {3,4,7}\{3,4,7\}, {3,5,7,13}\{3,5,7,13\}, {3,6,7,13,21}\{3,6,7,13,21\} and {3,4,5}\{3,4,5\}: yes, every integer above the stated value is represented. Not covered: exponents k≥2k\ge2 and every other tuple, and the first question. The site's commentary says the authors proved the conjecture for {3,4,7}\{3,4,7\}; the result covers only k=1k=1.

Depends on. No page of this wiki.

Acceptance. Refereed: Acta Arithmetica 77 (1996), no. 2, 133--138. The site's commentary (page last edited 1 December 2025) credits the {3,4,7}\{3,4,7\} case to this paper but labels the problem OPEN, so no reviewed evidence is listed. The formal-conjectures statement file for the problem states the {3,4,7}\{3,4,7\} case at k=1k=1 as erdos124.ne_zero_three_four_seven, marked research solved and credited to this paper, with no formal proof. The computations were not reconstructed in this corpus.