Inverse problem for critical restricted sumset pairs in Z_p

Let pp be prime, and let A,B⊆Z/pZA,B\subseteq\mathbb{Z}/p\mathbb{Z} be nonempty subsets with ∣A∣≠∣B∣|A|\ne|B|. Define the restricted sumset A∔B={a+b(modp):a∈A, b∈B, a≠b}A\dotplus B=\{a+b\pmod p:a\in A,\ b\in B,\ a\ne b\}. Classify all pairs (A,B)(A,B) satisfying ∣A∔B∣=min⁡{p,∣A∣+∣B∣−2}|A\dotplus B|=\min\{p,|A|+|B|-2\}, where the Alon--Nathanson--Ruzsa theorem guarantees the corresponding lower bound.

References

Progress summary

Refreshed
Claimed progress

A September 2026 paper reports that the conjecture fails at the boundary, while a narrower version works and the full question remains open.

The problem asks for an inverse classification of critical restricted sumset pairs in a prime cyclic group. The latest report says the conjecture fails when ∣A∣+∣B∣=p|A|+|B|=p, but a corrected result holds when ∣A∣+∣B∣≤p−1|A|+|B|\leq p-1.

Known results

For A,B⊆Z/pZA,B\subseteq\mathbb{Z}/p\mathbb{Z} with ∣A∣=∣B∣=k|A|=|B|=k, a 2024 paper claims that ∣A+˙B∣=2k−2|A\dot{+}B|=2k-2 implies A=BA=B for k≥5k\geq 5 and p>2k−2p>2k-2; Károlyi’s earlier work characterized the preceding equality case, with stated exceptions.

September 2026 counterexample

A report dated September 9, 2026 identifies a counterexample to the Liu–Qian conjecture at ∣A∣+∣B∣=p|A|+|B|=p and states a refined inverse theorem under ∣A∣+∣B∣≤p−1|A|+|B|\leq p-1. The unrestricted inverse problem therefore remains open.

Current status (as of September 2026): the boundary case ∣A∣+∣B∣=p|A|+|B|=p is claimed to be false and the restricted range ∣A∣+∣B∣≤p−1|A|+|B|\leq p-1 is claimed to be settled, but the unrestricted inverse problem remains open.

Sources

Solutions 0

No solutions have been posted yet.