Anti-diagonal core conjecture

Determine whether, for the reduced-word BFZ deep locus of the largest open double Bruhat cell G^{w_0,w_0}\subset SL_N, the anti-diagonal torus A_N^\circ is contained in every irreducible component, or at least every component invariant under anti-diagonal transpose, and prove whether A_N^\circ is exactly the universal intersection core of the irreducible components containing it.

Background

The anti-diagonal torus A_N\circ consists of determinant-one matrices supported on the anti-diagonal. The paper proves that it is contained in the reduced-word deep locus of the largest open cell and observes that in SL_3 it lies in the intersection of all displayed deep components.

The authors ask whether this behavior persists in higher rank: the anti-diagonal torus might be contained in all irreducible components, or might instead be a universal intersection core formed by the components that contain it.

References

This proposition leaves open another question: is the anti-diagonal torus itself an irreducible component, or is it always contained in larger deep components? The $SL_3$ calculation suggests the latter: the anti-diagonal support stratum is contained in several larger components, and we conjecture that in higher rank, the anti-diagonal may be better understood as a universal intersection core of the largest-cell deep locus.

On cluster deep loci in double Bruhat cells  (2608.27870 - Quan, 28 Aug 2026) in Section 8, subsection “The anti-diagonal stratum,” especially Conjecture 8.2

For $G=SL_N$, the anti-diagonal torus $A_N\circ$ is contained in every irreducible components of $\mathcal D{w_0,w_0}$, or the ones that are invariant under anti-diagonal transpose.

More generally, $A_N\circ$ should be exactly a universal intersection core, which means its is the intersection of all irreducible components containing it.

On cluster deep loci in double Bruhat cells  (2608.27870 - Quan, 28 Aug 2026) in Conjecture “Anti-diagonal core conjectures,” Section 8