← All problems

Prasad’s conjecture on quaternionic symplectic models

Let FF be a non-archimedean local field of characteristic different from 22, let DD be a quaternion division algebra over FF, and let G=GL⁡n(D)G=\operatorname{GL}_n(D). Write H=Sp⁡n(D)={g∈GL⁡n(D):t ⁣g‾g=1}H=\operatorname{Sp}_n(D)=\{g\in \operatorname{GL}_n(D):{}^t\!\overline{g}g=1\}, where x‾\overline{\phantom{x}} is the canonical involution of DD. For every irreducible admissible representation π\pi of GG, one has Hom⁡H(π,1)≠0\operatorname{Hom}_H(\pi,\mathbf{1})\neq 0 if and only if the Jacquet--Langlands transfer JL⁡(π)\operatorname{JL}(\pi) to GL⁡2n(F)\operatorname{GL}_{2n}(F) is not cuspidal.

References

Source: arXiv

Progress summary

Refreshed
Open · progress reported

New preprints claim substantial advances for the cuspidal and tempered cases, but the full conjecture remains unsettled and the claims are unverified.

The conjecture concerns when representations of quaternionic general linear groups admit symplectic periods, with a predicted characterization via Jacquet–Langlands transfer. The retrieved literature records substantial results for cuspidal, discrete-series, and tempered representations, but not a verified resolution of the entire conjecture.

Known results

  • Verma proved the forward implication in characteristic zero; Sécherre and Stevens state an extension to arbitrary characteristic.
  • Sécherre and Stevens, submitted March 12, 2025: the cuspidal case is stated and proved, including the criterion that the Jacquet–Langlands transfer is non-cuspidal.
  • Results for n=3n=3 and n=4n=4 classify the predicted distinguished unitary representations in the cases treated, while some supercuspidal cases remained open.

October 2026 preprint developments

Fei Chen and Nadir Matringe claim an algebraic treatment removing unimodularity assumptions, allowing positive odd characteristic and twisted periods, and proving the cuspidal case. Another October 2026 preprint claims the revised conjecture for tempered representations. Both are unrefereed and lack independent verification in the retrieved sources.

Current status (as of October 2026): the cuspidal and some discrete-series cases are claimed in preprints, and a tempered classification is also claimed, but the full conjecture has not been independently verified or established here.

Sources

Solutions 0

No solutions have been posted yet.