Three-translate linear stability conjecture for finite subsets of the integers

Let AA and BB be finite subsets of Z\mathbb{Z} of equal size nn. For b1,b2,b3Bb_1,b_2,b_3\in B, write A+{b1,b2,b3}={a+bi:aA,i{1,2,3}}A+\{b_1,b_2,b_3\}=\{a+b_i:a\in A,\,i\in\{1,2,3\}\}. For finite sets X,YX,Y, write XΔYX\mathbin{\Delta}Y for their symmetric difference.

Three-translate linear stability conjecture. For every δ>0\delta>0 there exists ε>0\varepsilon>0 such that, if

A+{b1,b2,b3}(2+ε)n1|A+\{b_1,b_2,b_3\}|\leqslant (2+\varepsilon)n-1

for every b1,b2,b3Bb_1,b_2,b_3\in B, then there exist arithmetic progressions P,QP,Q in Z\mathbb{Z} with the same common difference such that

BQandAΔP, BΔQδn.B\subset Q\qquad\text{and}\qquad |A\mathbin{\Delta}P|,\ |B\mathbin{\Delta}Q|\leqslant\delta n.

This is the proposed three-translate analogue of the paper's inverse theorem for four translates. The source presents this weaker stability form as conjectural, while noting that a stronger version is also expected.

Sources & referencesView supporting material

Primary source

Bela Bollobas, Imre Leader and Marius Tiba, “A strengthening of Freiman's 3k-4 theorem”, arXiv:2204.09816 (2022).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.