What the literature proves

Which known results are closest to a combinatorial generation criterion?

1 · Scope of the comparison

This review checks HKP’s algebraic input, Chen–Yi’s two-basis comparison, and Shtotland’s orbit-module presentation. These are closely related ingredients; none is quoted as a theorem asserting that one length-zero element forces generation from a prescribed basic vector for every affine spherical variety.

Table 1. What each source contributes.
Source	Input obtained	What still needs proof for X
[HKP, §7.1, p. 19]	Length-zero group algebra	Which supports its translates reach
[CY, §8.2.3 and Lemma 73, pp. 42–43]	Standard/IC basis comparison	Cyclicity of the chosen basic vector
[Sht, Definitions 8.1, 8.4; Theorem 8.5, pp. 31–32]	Weighted orbit description of a symmetric-space module	A criterion for generation by the chosen basic vector

2 · Symmetric spaces and length-zero permutations

· The setting of Shtotland Definition

Let $\Loc$ be a p-adic field with residue characteristic different from 2, $\Grp$ a connected split reductive group, and $\theta:\Grp\to\Grp$ an algebraic involution. Set $H=\Grp^\theta$ and use the source’s right homogeneous space:

$$X(\Loc)=H(\Loc)\backslash\Grp(\Loc),\qquad \mathscr O_X=H(\Loc)\backslash\Grp(\Loc)/\Iw.$$

The orbit-indicator functions $\mathbf1_x$, indexed by $x\in\mathscr O_X$, form a basis of the Iwahori-fixed Schwartz space. These orbits are compact open because the space is homogeneous.

· Ω permutes orbit indicators Proved (cited)

$$\mathbf1_x T_\omega=\mathbf1_{x\times\omega}\qquad(x\in\mathscr O_X,\ \omega\in\Om).$$[Sht, Proposition 4.1, p. 17]

Proof

Right convolution has support in one translated orbit because $\omega$ normalizes $\Iw$: $$H(\Loc)x\Iw\omega\Iw=H(\Loc)x\omega\Iw.$$
On that orbit the integrand is one, and the double coset has volume one; elsewhere it is zero: $$\mathbf1_xT_\omega(y)=\mathbf1_{H(\Loc)x\omega\Iw}(y).$$

Consequently, if Ω is transitive on the orbit set, an orbit indicator is a cyclic vector already for $\mathbb C[\Om]$. If the basic function is a sum of orbit indicators, cyclicity requires a separate check.

3 · More than one permutation

In the source’s Iwahori hypergraph $\mathscr G_X$, vertices are I-orbits and hyperedges record rank-one parahoric incidences; a length function and coefficients determine reflection actions. With formal Hecke parameter $u$, the construction specializes to the actual module:

$$\alpha_q:M_u(\mathscr G_X)/(u-q)\xrightarrow{\sim}C_c^\infty(X(\Loc))^{\Iw}.$$ $$\alpha_q(v_x)=\mathbf1_x.$$[Sht, Definitions 8.1, 8.4; Theorem 8.5, pp. 31–32]

The coefficient data matter: a reflection can produce a combination of several orbit indicators. This supplies a framework for testing cyclicity by linear algebra; it does not replace that test with the existence of τ.

The finite-generation theorem gives a finite set of module generators under the source’s hypotheses.

[Sht, Theorem 5.9, pp. 22–23]

It does not single out $\mathbf1_{X(\Ring)}$. In affine space, the stronger one-vector statement comes from the special chain of smooth closures computed on the basis page.

4 · The formal research question

Question Fix a class of smooth affine spherical G-varieties with an integral model and a finite-level orbit presentation. Can one characterize, from its weighted orbit data and the normalizer action, when

$$\mathbb C[\Om]*\mathbf1_{X(\Ring)}=C_c^\infty(X(\Loc))^{\Iw}?$$

If Ω-translates do not span, can the full weighted Hecke action determine whether

$$\Hecke*\mathbf1_{X(\Ring)}=C_c^\infty(X(\Loc))^{\Iw}?$$

The first equality is stronger. The finite-level triangular criterion proves a sufficient test; the sheaf version additionally asks for all simple local systems and bounded finite-level generation.