ProbXiv
Explorersign in
Problem archiveProblem record

Statement

Let G\mathbf G be an algebraic group over Fˉq\bar{\mathbb F}_q, let FF be a Frobenius map, let G=GFG=\mathbf G^F, and let WW be the FF-split Weyl group. Let w∈Ww\in W be a good element of order dd, let XwX_w be the associated Deligne–Lusztig variety, and fix a prime ℓ\ell not dividing qq. Write Hci(Xw,Qˉℓ)H^i_c(X_w,\bar{\mathbb Q}_\ell) for its étale cohomology, and let H(G,F,W,w)\mathcal H(\mathbf G,F,W,w) be the image of QˉℓBw\bar{\mathbb Q}_\ell B_w in the graded endomorphism algebra of ⨁i≥0Hci(Xw,Qˉℓ)\bigoplus_{i\geq0}H^i_c(X_w,\bar{\mathbb Q}_\ell). Broué–Malle–Michel's conjecture. Suppose that ww is a good element of order dd. Then: (1) if i≠ji\ne j, the QˉℓGF\bar{\mathbb Q}_\ell\mathbf G^F-modules Hci(Xw,Qˉℓ)H^i_c(X_w,\bar{\mathbb Q}_\ell) and Hcj(Xw,Qˉℓ)H^j_c(X_w,\bar{\mathbb Q}_\ell) have no irreducible constituents in common; (2) there is a dd-cyclotomic Hecke algebra Hx(CW(w))\mathscr H_x(C_W(w)) such that

H(G,F,W,w)≅HQˉℓ,q(CW(w))≅End⁡QˉℓGF(⨁i≥0Hci(Xw,Qˉℓ));\mathcal H(\mathbf G,F,W,w)\cong\mathscr H_{\bar{\mathbb Q}_\ell,q}(C_W(w))\cong\operatorname{End}_{\bar{\mathbb Q}_\ell\mathbf G^F}\left(\bigoplus_{i\geq0}H^i_c(X_w,\bar{\mathbb Q}_\ell)\right);

(3) there is a one-to-one correspondence χ⟷χq\chi\longleftrightarrow\chi_q between irreducible representations of CW(w)C_W(w) and irreducible constituents of ⨁i≥0Hci(Xw,Qˉℓ)\bigoplus_{i\geq0}H^i_c(X_w,\bar{\mathbb Q}_\ell), and for each irreducible character χ\chi there is a polynomial Dχ(x)D_\chi(x), depending only on χ\chi, such that deg⁡(χq)=Dχ(q)\deg(\chi_q)=D_\chi(q). The conjecture proposes a uniform description of the cohomology and its endomorphism algebra in terms of cyclotomic Hecke algebras; the supplied text does not state whether it has been resolved.

Record

Comments

Nothing has been published against this problem yet, and nothing has been said about it.

Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.

Sign in

Solve with an agent

Open the statement in a chat, with the problem and the ground rules already written into the prompt.

This opens a third-party site. Nothing is posted back to ProbXiv and nothing you write there is recorded here — what a model gives you is an attempt, which a person still has to check.