← All problems

Linearity Conjecture for minimal blocking sets

Let q=phq=p^h be a prime power, and let BB be a small minimal blocking set with respect to lines in PG⁡(2,q)\operatorname{PG}(2,q), meaning that BB meets every line, no proper subset of BB has this property, and ∣B∣<3(q+1)2|B|<\frac{3(q+1)}{2}. Let pep^e be the exponent of BB, defined as the largest power of pp such that every line meets BB in a number of points congruent to 1(modpe)1\pmod{p^e}. Then BB is an Fpe\mathbb{F}_{p^e}-linear point set: there exists an Fpe\mathbb{F}_{p^e}-vector subspace U≤Fq3U\leq\mathbb{F}_q^3 such that B={⟨u⟩Fq:u∈U∖{0}}B=\{\langle u\rangle_{\mathbb{F}_q}:u\in U\setminus\{0\}\}.

References

Source: arXiv

Progress summary

Refreshed
Open · progress reported

An unrefereed paper claims to settle the conjecture in every case except exponents three and four, which remain open.

The Linearity Conjecture concerns whether small minimal blocking sets in projective spaces are linear over a subfield. Earlier work established several special cases, while the general statement remained open.

Known results

  • A 2012 paper records cases t=1t=1, t=2t=2, and t=3t=3.
  • Rédei-type blocking sets were known to be linear.
  • Blocking sets spanning a tktk-dimensional space were known to be linear.
  • A related 2012 result proves linearity for sets spanning dimension t/e−1t/e-1 under hypotheses including pe≥7p^e\ge 7.

October 2026 claimed advance

Peter Sziklai’s preprint The linearity Conjecture claims the conjecture for pe=2p^e=2 and for every pe≥5p^e\ge 5, leaving only pe=3p^e=3 and pe=4p^e=4. This is an unrefereed claim, and the supplied scan found no independent verification or assessment.

Current status (as of October 2026): Earlier special cases are established, while Sziklai’s claimed extension to pe=2p^e=2 and pe≥5p^e\ge 5 is unverified and the cases pe=3p^e=3 and pe=4p^e=4 remain open in the stated result.

Sources

Solutions 0

No solutions have been posted yet.