← All problems

Luo’s conjecture on separating elements of 3-manifold groups using finite PGL₂ over rings

For every compact 33-manifold MM and every nontrivial element g∈π1(M)g\in \pi_1(M), there exist a finite commutative ring RR with identity and a homomorphism ρ:π1(M)→PGL2(R)\rho:\pi_1(M)\to \mathrm{PGL}_2(R) such that ρ(g)≠1\rho(g)\ne 1. Equivalently, every compact 33-manifold group is residually finite through groups of the form PGL2(R)\mathrm{PGL}_2(R), with RR a finite commutative ring with identity.

References

Progress summary

Refreshed
Open · progress reported

The conjecture is false in some formulations, while a 2026 correction repairs two variants without settling the general question.

Luo’s conjecture asks whether every nontrivial element of the fundamental group of a compact 33-manifold can be separated in some finite group of the form PGL(2,R)\mathrm{PGL}(2,R). The literature gives counterexamples to broad formulations, but the remaining scope and related representation variants are not completely settled.

Known results

  • Friedl, Gill, and Tillmann, 2017: reformulated the conjecture via representations over commutative rings and recorded counterexamples from earlier work.
  • Gill, 2026: for geometrizable manifolds, the conjecture holds for six of Thurston’s eight geometries.
  • Gill, 2026: π1(Σ(2,3,5))\pi_1(\Sigma(2,3,5)) is not residually PGL2\mathrm{PGL}_2-finite or PSL2\mathrm{PSL}_2-finite.
  • Gill, 2026: π1(Σ(2,3,7))\pi_1(\Sigma(2,3,7)) is not residually finite through any of GL2\mathrm{GL}_2, SL2\mathrm{SL}_2, PGL2\mathrm{PGL}_2, or PSL2\mathrm{PSL}_2 over finite commutative rings.

October 2026 correction

Montek Singh Gill reports that the earlier argument established only the SL2\mathrm{SL}_2 and GL2\mathrm{GL}_2 variants; an amended argument restores the PSL2\mathrm{PSL}_2 and PGL2\mathrm{PGL}_2 cases. This unrefereed correction advances the variant results but does not settle Luo’s conjecture in general.

Current status (as of November 2026): Specific manifolds and variants are settled, but the general conjecture remains open; the 2026 correction is an unverified claimed advance.

Sources

Solutions 0

No solutions have been posted yet.