← Back to arXiv
arXivNumber TheoryarXiv:2609.38231

Strong Artin Conjecture for Generalized Octahedral Representations in $\mathrm{GL}_3$

The Artin conjecture is a century-old problem in number theory that predicts certain functions called L-functions, built from symmetry data of number fields, should be "entire," meaning they behave as nicely as possible across the complex plane. These L-functions encode deep arithmetic information, and proving they are entire amounts to showing that the symmetry representations they come from correspond to well-understood objects in a different mathematical framework called automorphic forms. Proving this correspondence is often extremely difficult and is only known in special cases.

The paper focuses on three-dimensional representations, meaning symmetry actions on three-dimensional space, where the group of symmetries is solvable (roughly, built up from simple abelian pieces in a structured way). The authors tackle the last and hardest remaining solvable case in three dimensions: representations whose symmetry group has a specific structure involving a combination of cyclic and special linear groups. To handle this case, they develop new machinery around "base change" and "automorphic induction," which are technical tools for transferring automorphic forms between different number fields. A particular challenge here is that the relevant field extensions lack intermediate fields, making standard approaches unavailable.

By resolving this final three-dimensional solvable case, the authors complete the full picture for solvable Artin representations in three dimensions. As a bonus, their methods also settle the strong Artin conjecture for all six-dimensional primitive solvable representations, since those can be related back to the three-dimensional case through a natural lifting process. The result closes a long-standing gap in the classification of solvable Artin representations and provides new tools that may be useful for attacking non-solvable cases in the future.

Read original →