§6 · Agent 6 · Audit: integral models · 2026-10-18T04:01:10Z
Audit 6.7 · Audit of Sakellaridis's scheme \(Y\) and the complement clause (claims A2-1, A2-8)
proved audit: correct ★ key step
Second audit: correct. Checked against pp. 7 and 9 of arXiv:0905.4244v3: X-ring and Y are defined with the inner complement in the generic fibre and the outer one in the O-closure, and Audit 6.7's points (a)-(c) correctly show this is the only satisfiable reading (complements in X are impossible whenever (YP_alpha)_F misses the open orbit; for the open saturations the two readings coincide since the closure is X). The identification Y^can = Y_x (Y_x open in Z_x by flatness; boundary closures with special fibres the unions over Sigma_v, Sigma_v in Sigma_x minus x; equal reduced open subschemes) and the complement clause for type U (E = Sigma_u minus {u,u'} is Sigma-stable by the dimension argument) re-derive correctly. The remark that the 45 orbits have Z[1/2]-points, so the characteristic-0 citation [Sa06, 3.6.1] is not needed, is right.
The text of Axiom 2.2.1Sakellaridis's Axiom 2.2.1, relabelled (A3a): rank-one models over O. forces the reading used by Strategy 2.1 (complements inside closures), the reading with complements in \(X\) being unsatisfiable for every non-open saturation; with that reading Theorem 2.8 correctly identifies Sakellaridis's \(Y\) with the orbit scheme \(Y_x\) and proves the complement clause for all 270 pairs. Outcome: holds.
Audit 6.7.
- On p. 9 of arXiv:0905.4244v3 (with \(k=F\)), the scheme structure is \(Y^{\rm can}=\overline Y\smallsetminus\overline{(\overline{Y_F}\smallsetminus Y_F)}\), closures in \(X_{\mathcal O}\) and the inner complement in \(X_F\); the clause "the complement of \(YP_\alpha\) is the closure of the complement of \((YP_\alpha)_F\)" means \(P_\alpha Y^{\rm can}=(P_\alpha Y_F)^{\rm can}\). Read with complements in \(X\), the clause fails for every pair with \((YP_\alpha)_F\) not open, in any model; for the three open saturations \(\mathring X\), \(P_{\alpha_1}\mathring X\), \(P_{\alpha_6}\mathring X\) both readings hold for \(\mathcal M_\theta\).
- For every \(x\in\mathcal C\): \(Y^{\rm can}(Y_{x,F})=Y_x\) as subschemes of \(X_{\mathcal O}\); for every simple \(\alpha\), \(P_\alpha Y_x=(P_\alpha Y_{x,F})^{\rm can}\), and \(P_\alpha Y_x/R(P_\alpha)\) is \(\operatorname{Spec}\mathcal O\) (90 pairs) or \(\mathbb P^1_{\mathcal O}\) (180 pairs). Outcome: holds.
- Text. Axiom 2.2.1Sakellaridis's Axiom 2.2.1, relabelled (A3a): rank-one models over O. (p. 9): "For every \(B_k\)-orbit \(Y_k\subset X_k\) of maximal rank and every simple root \(\alpha\), the o-scheme \(YP_\alpha/R(P_\alpha)\) is isomorphic to one of: \(\mathrm{PGL}_2\backslash\mathrm{PGL}_2\), \(SU\backslash\mathrm{PGL}_2\) (where \(S\subset N(U)\)), \(T\backslash\mathrm{PGL}_2\) … or \(N(T)\backslash\mathrm{PGL}_2\). Moreover, the complement of \(YP_\alpha\) is the closure of the complement of \((YP_\alpha)_k\)." Then: "Y is the complement of the closure of the complement of \(Y_k\) in its closure."
- Which complements. (a) If the inner complement were taken in the \(\mathcal O\)-closure \(\overline Y\), it would contain the special fibre of \(\overline Y\), and \(Y\) would be \(Y_F\) with empty special fibre, contradicting Sakellaridis's next sentence (Y smooth and surjective over o). (b) If the outer complement were taken in \(X\), \(Y\) would be open in \(X\). So \(Y^{\rm can}\) is as in Strategy 2.1. (c) For the clause: if \((YP_\alpha)_F\) is not open it does not contain \(\mathring X_F\) (a \(B\)-stable set meeting \(\mathring X_F\) contains it, and \(P_\alpha\mathring X_F\) is open), so \(X_F\smallsetminus(YP_\alpha)_F\supseteq\mathring X_F\) has closure \(X\), and the \(X\)-reading would force \(YP_\alpha=\emptyset\). The closure reading is the only satisfiable one. Strategy 2.1 reads it so.
- \(Y^{\rm can}=Y_x\). \(Y_x\) is locally closed and \(\mathcal O\)-flat, so its topological closure is \(\overline{Y_{x,F}}=Z_x\) and it is open in \(Z_x\). \(\partial_F=\bigcup_{v\in\Sigma_x\smallsetminus\{x\}}Y_{v,F}\) has closure \(\bigcup Z_v\). By Lemma 2.7 item 3 (audited in Audit 6.6) \(Z_v\cap X_\kappa=\bigcup_{\Sigma_v}Y_{u,\kappa}\), and \(\Sigma_v\subseteq\Sigma_x\smallsetminus\{x\}\) (transitivity, Audit 6.1 item 6; \(x\notin\Sigma_v\) by dimension). So \(Y^{\rm can}\cap X_\kappa=\overline{Y_{x,\kappa}}\smallsetminus\bigcup_{\Sigma_x\smallsetminus\{x\}}Y_{v,\kappa}=Y_{x,\kappa}\), and \(Y^{\rm can}_F=Y_{x,F}\). Both are reduced open subschemes of the reduced \(Z_x\) with the same support, hence equal.
- Complement clause. Type G: \(P_\alpha Y_x=Y_x\) (Audit 6.5 step 2), and the clause is item 2 for \(x\). Type U: with \(u\) the open member of \(\{x,s_\alpha xs_\alpha\}\) and \(u'=s_\alpha us_\alpha\), \(\mathcal P=P_\alpha Y_x\) is smooth, flat and locally closed, so open in its closure \(Z_u\); \(E=\Sigma_u\smallsetminus\{u,u'\}\) is stable under \(v\mapsto\Sigma_v\) (if \(v\in E\) and \(u'\in\Sigma_v\) then \(\dim Y_v\ge\dim Y_{u'}=\dim Y_u-1\), forcing \(v=u'\)); so, as in step 3, \((P_\alpha Y_{x,F})^{\rm can}\) has special fibre \(Y_{u,\kappa}\sqcup Y_{u',\kappa}=|\mathcal P_\kappa|\) and equals \(\mathcal P\). The quotients are those of Audit 6.5.
- Open saturations, \(X\)-reading. \(X_{\mathcal O}\) is \(\mathcal O\)-smooth with irreducible fibres, so \(Z_{w_L}=X_{\mathcal O}\) and \(\mathring X=Y_{w_L}\) (consistent with S4-4). \(X\smallsetminus Y_{w_L}=\bigcup_{v\ne w_L}Y_v\) and \(\overline{X_F\smallsetminus\mathring X_F}=\bigcup_{v\ne w_L}Z_v\) have the same special fibre by Lemma 2.7; likewise for \(P_{\alpha_1}\mathring X\), \(P_{\alpha_6}\mathring X\) with \(E=\mathcal C\smallsetminus\{w_L,s_\alpha w_Ls_\alpha\}\), as Theorem 2.8 step 7 says.
- "\(B_F\)-orbit" over a non-closed \(F\). The 45 orbits are base changes of subschemes over \(\mathbb Z[1/2]\) with \(\mathbb Z[1/2]\)-points, so the \(B_F\)-orbits are unambiguous in both characteristics; Sakellaridis's citation of [Sa06, Prop. 3.6.1] (rationality of orbits, characteristic 0) is not needed.
Outcome: holds.
Builds on
- Strategy 2.1 · Strategy: Axiom 2.2.1Sakellaridis's Axiom 2.2.1, relabelled (A3a): rank-one models over O. for \(\mathcal M_\theta\) once over \(\mathbb Z[1/2]\)
- Theorem 2.8 · Theorem: Axiom 2.2.1Sakellaridis's Axiom 2.2.1, relabelled (A3a): rank-one models over O. holds for \(\mathcal M_\theta\) over every DVR with \(2\) invertible
Built on by
- Summary 6.9 · Verdict: (A3a)(A3a) Rank-one models over O (Sakellaridis's Axiom 2.2.1): for every Borel orbit Y of maximal rank and every simple root α, the O-scheme YP_α/R(P_α) is one of the four standard PGL2-spaces over O, with his scheme structure on Y, and YP_α is the closure of its generic fibre minus the closure of the generic boundary. holds for \(\mathcal M_\theta\) over every DVR with \(2\) invertible
Sources
- sakellaridis-v3.pdf, Axiom 2.2.1Sakellaridis's Axiom 2.2.1, relabelled (A3a): rank-one models over O. and the scheme structure on Y, p. 9
- S4.json: S4-4 (open orbit)