Equivariant cohomology and fixed points of smooth Calogero-Moser spaces
C´edric Bonnaf´e
CNRS (UMR 5149) - Universit´e de Montpellier
Banff, Canada, October 2017
Set-up
Set-up
dimC(V) =n<∞
W <GLC(V), |W|<∞
Set-up
dimC(V) =n<∞
W <GLC(V), |W|<∞ cod(w) = codimC(Vw)
cod(ww′)6cod(w) +cod(w′)
Set-up
dimC(V) =n<∞
W <GLC(V), |W|<∞ cod(w) = codimC(Vw)
cod(ww′)6cod(w) +cod(w′) Ref(W) =cod−1(1).
Hypothesis: W =hRef(W)i
Set-up
dimC(V) =n<∞
W <GLC(V), |W|<∞ cod(w) = codimC(Vw)
cod(ww′)6cod(w) +cod(w′) Ref(W) =cod−1(1).
Hypothesis: W =hRef(W)i
Set-up (continued)
Set-up (continued)
Fi(CW) =L
cod(w)6iCw
Fi(CW)·Fj(CW)⊂Fi+j(CW)
Set-up (continued)
Fi(CW) =L
cod(w)6iCw
Fi(CW)·Fj(CW)⊂Fi+j(CW)
F0(CW) = C⊂F1(CW)⊂F2(CW)⊂ · · · ⊂Fn(CW) =CW is a filtration ofCW.
Set-up (continued)
Fi(CW) =L
cod(w)6iCw
Fi(CW)·Fj(CW)⊂Fi+j(CW)
F0(CW) = C⊂F1(CW)⊂F2(CW)⊂ · · · ⊂Fn(CW) =CW is a filtration ofCW.
If Ais a subalgebra of CW, we define Fi(A) =A∩Fi(CW)
Set-up (continued)
Fi(CW) =L
cod(w)6iCw
Fi(CW)·Fj(CW)⊂Fi+j(CW)
F0(CW) = C⊂F1(CW)⊂F2(CW)⊂ · · · ⊂Fn(CW) =CW is a filtration ofCW.
If Ais a subalgebra of CW, we define Fi(A) =A∩Fi(CW) ReesF(A) =L
i>0hiFi(A)⊂C[h]⊗A
Set-up (continued)
Fi(CW) =L
cod(w)6iCw
Fi(CW)·Fj(CW)⊂Fi+j(CW)
F0(CW) = C⊂F1(CW)⊂F2(CW)⊂ · · · ⊂Fn(CW) =CW is a filtration ofCW.
If Ais a subalgebra of CW, we define Fi(A) =A∩Fi(CW) ReesF(A) =L
i>0hiFi(A)⊂C[h]⊗A C[h]/hh−ξi ⊗ReesF(A)≃
A if ξ6=0, GradF(A) if ξ=0.
Example: W = S
n⊂ GL
n( C )
Example: W = S
n⊂ GL
n( C )
Part(n) ←→∼ Irr(Sn) λ 7−→ χλ
Example: W = S
n⊂ GL
n( C )
Part(n) ←→∼ Irr(Sn) λ 7−→ χλ
λ7−→
γd(λ)∈Part →its d-core λ[d]∈Partd →its d-quotient
Example: W = S
n⊂ GL
n( C )
Part(n) ←→∼ Irr(Sn) λ 7−→ χλ
λ7−→
γd(λ)∈Part →its d-core λ[d]∈Partd →its d-quotient
|λ|=|γd(λ)|+d |λ[d]|.
Example: W = S
n⊂ GL
n( C )
Part(n) ←→∼ Irr(Sn) λ 7−→ χλ
λ7−→
γd(λ)∈Part →its d-core λ[d]∈Partd →its d-quotient
|λ|=|γd(λ)|+d |λ[d]|.
Theorem (Fong-Srinivasan, 1982)
Assume thato(q mod ℓ) =d. Two unipotent characters ρλ and ρµ of GLn(Fq)lie in the same ℓ-block if and only if γd(λ) = γd(µ).
Example: W = S
n⊂ GL
n( C ) (continued)
Example: W = S
n⊂ GL
n( C ) (continued)
Fix a d-coreγ such thatr = n−|γ|
d ∈N
Example: W = S
n⊂ GL
n( C ) (continued)
Fix a d-coreγ such thatr = n−|γ|
d ∈N Sr ⋉µrd ≃G(d,1,r)⊂GLr(C)
Example: W = S
n⊂ GL
n( C ) (continued)
Fix a d-coreγ such thatr = n−|γ|
d ∈N Sr ⋉µrd ≃G(d,1,r)⊂GLr(C)
Partd(r) ←→∼ Irr(G(d,1,r)) µ 7−→ χ(d)µ
Example: W = S
n⊂ GL
n( C ) (continued)
Fix a d-coreγ such thatr = n−|γ|
d ∈N Sr ⋉µrd ≃G(d,1,r)⊂GLr(C)
Partd(r) ←→∼ Irr(G(d,1,r)) µ 7−→ χ(d)µ
Let ϕγ : Z(CSn) −→ Z(CG(d,1,r))
eχλ 7−→
eχ(d)
λ[d]
if γd(λ) =γ, 0 if γd(λ)6=γ,
Example: W = S
n⊂ GL
n( C ) (continued)
Fix a d-coreγ such thatr = n−|γ|
d ∈N Sr ⋉µrd ≃G(d,1,r)⊂GLr(C)
Partd(r) ←→∼ Irr(G(d,1,r)) µ 7−→ χ(d)µ
Let ϕγ : Z(CSn) −→ Z(CG(d,1,r))
eχλ 7−→
eχ(d)
λ[d]
if γd(λ) =γ, 0 if γd(λ)6=γ,
Theorem (B.-Maksimau-Shan, 2017)
ϕγ(Fi(Z(CSn))⊂Fi(Z(CG(d,1,r)))
Example: W = S
n⊂ GL
n( C ) (continued)
Fix a d-coreγ such thatr = n−|γ|
d ∈N Sr ⋉µrd ≃G(d,1,r)⊂GLr(C)
Partd(r) ←→∼ Irr(G(d,1,r)) µ 7−→ χ(d)µ
Let ϕγ : Z(CSn) −→ Z(CG(d,1,r))
eχλ 7−→
eχ(d)
λ[d]
if γd(λ) =γ, 0 if γd(λ)6=γ,
Theorem (B.-Maksimau-Shan, 2017)
ϕγ(Fi(Z(CSn))⊂Fi(Z(CG(d,1,r)))
(i.e. (IdC[h]⊗ϕγ)(ReesFZ(CSn))⊂ ReesFZ(CG(d,1,r)) ).
Calogero-Moser spaces
Calogero-Moser spaces
c :Ref(W)/∼ −→ C
Hc =C[V]⊗CW ⊗C[V∗] (as a vector space)
∀y ∈V, ∀x ∈V∗, [y,x] = X
s∈Ref(W)
cshy,s(x) −xis
Calogero-Moser spaces
c :Ref(W)/∼ −→ C
Hc =C[V]⊗CW ⊗C[V∗] (as a vector space)
∀y ∈V, ∀x ∈V∗, [y,x] = X
s∈Ref(W)
cshy,s(x) −xis
Calogero-Moser spaces
c :Ref(W)/∼ −→ C
Hc =C[V]⊗CW ⊗C[V∗] (as a vector space)
∀y ∈V, ∀x ∈V∗, [y,x] = X
s∈Ref(W)
cshy,s(x) −xis
Zc =Z(Hc)
Zc =Specmax(Zc) (Calogero-Moser space)
Calogero-Moser spaces
c :Ref(W)/∼ −→ C
Hc =C[V]⊗CW ⊗C[V∗] (as a vector space)
∀y ∈V, ∀x ∈V∗, [y,x] = X
s∈Ref(W)
cshy,s(x) −xis
Zc =Z(Hc)
Zc =Specmax(Zc) (Calogero-Moser space) Example - Assume that c =0.
Calogero-Moser spaces
c :Ref(W)/∼ −→ C
Hc =C[V]⊗CW ⊗C[V∗] (as a vector space)
∀y ∈V, ∀x ∈V∗, [y,x] = X
s∈Ref(W)
cshy,s(x) −xis
Zc =Z(Hc)
Zc =Specmax(Zc) (Calogero-Moser space) Example - Assume that c =0.
H0 =C[V ×V∗]⋊W
Calogero-Moser spaces
c :Ref(W)/∼ −→ C
Hc =C[V]⊗CW ⊗C[V∗] (as a vector space)
∀y ∈V, ∀x ∈V∗, [y,x] = X
s∈Ref(W)
cshy,s(x) −xis
Zc =Z(Hc)
Zc =Specmax(Zc) (Calogero-Moser space)
Example - Assume that c =0.
H0 =C[V ×V∗]⋊W Z0 =C[V ×V∗]W
Calogero-Moser spaces
c :Ref(W)/∼ −→ C
Hc =C[V]⊗CW ⊗C[V∗] (as a vector space)
∀y ∈V, ∀x ∈V∗, [y,x] = X
s∈Ref(W)
cshy,s(x) −xis
Zc =Z(Hc)
Zc =Specmax(Zc) (Calogero-Moser space)
Example - Assume that c =0.
H0 =C[V ×V∗]⋊W Z0 =C[V ×V∗]W Z0 = (V ×V∗)/W.
Calogero-Moser space (continued)
Calogero-Moser space (continued)
Theorem (Etingof-Ginzburg, 2002)
(a) Zc is integral and integrally closed
Calogero-Moser space (continued)
Theorem (Etingof-Ginzburg, 2002)
(a) Zc is integral and integrally closed (i.e. Zc is irreducible and normal)
Calogero-Moser space (continued)
Theorem (Etingof-Ginzburg, 2002)
(a) Zc is integral and integrally closed (i.e. Zc is irreducible and normal)
(b) If Zc is smooth, then
Hc-mod −→ Zc-mod M 7−→ eM
is a Morita equivalence (here, e = |W1|P
w∈W w).
Calogero-Moser space (continued)
Theorem (Etingof-Ginzburg, 2002)
(a) Zc is integral and integrally closed (i.e. Zc is irreducible and normal)
(b) If Zc is smooth, then
Hc-mod −→ Zc-mod
M 7−→ eM = (eHc)⊗Hc M is a Morita equivalence (here, e = |W1|P
w∈W w).
Calogero-Moser space (continued)
Theorem (Etingof-Ginzburg, 2002)
(a) Zc is integral and integrally closed (i.e. Zc is irreducible and normal)
(b) If Zc is smooth, then
Hc-mod −→ Zc-mod
M 7−→ eM = (eHc)⊗Hc M is a Morita equivalence (here, e = |W1|P
w∈W w).
(c) If Zc is smooth, then
H2i+1(Zc) =0 H2•(Zc)|{z}≃
C-alg
GradFZ(CW)
Calogero-Moser space (continued)
Calogero-Moser space (continued)
Hc isZ-graded with
deg(V) = −1, deg(V∗) =1 and deg(W) =0.
Calogero-Moser space (continued)
Hc isZ-graded with
deg(V) = −1, deg(V∗) =1 and deg(W) =0.
⇒ Zc is Z-graded (i.e. admits a C×-action)
Calogero-Moser space (continued)
Hc isZ-graded with
deg(V) = −1, deg(V∗) =1 and deg(W) =0.
⇒ Zc is Z-graded (i.e. admits a C×-action)
⇒ Zc admits aC×-action.
Calogero-Moser space (continued)
Hc isZ-graded with
deg(V) = −1, deg(V∗) =1 and deg(W) =0.
⇒ Zc is Z-graded (i.e. admits a C×-action)
⇒ Zc admits aC×-action.
Theorem (Gordon, 2003)
IfZc issmooth, then
ZC×
c
←→∼ Irr(W)
Equivariant cohomology
Equivariant cohomology
If X is an algebraic variety endowed with a C×-action, one can define its equivariant cohomologyH•C×(X).
Equivariant cohomology
If X is an algebraic variety endowed with a C×-action, one can define its equivariant cohomologyH•C×(X).
H•C×(pt)≃C[h], with deg(h) =2.
Equivariant cohomology
If X is an algebraic variety endowed with a C×-action, one can define its equivariant cohomologyH•C×(X).
H•C×(pt)≃C[h], with deg(h) =2.
⇒ H•C×(X) is a C[h]-algebra.
Equivariant cohomology
If X is an algebraic variety endowed with a C×-action, one can define its equivariant cohomologyH•C×(X).
H•C×(pt)≃C[h], with deg(h) =2.
⇒ H•C×(X) is a C[h]-algebra.
Theorem (folklore)
Assume thatH2i+1(X) =0 for all i. Then:
Equivariant cohomology
If X is an algebraic variety endowed with a C×-action, one can define its equivariant cohomologyH•C×(X).
H•C×(pt)≃C[h], with deg(h) =2.
⇒ H•C×(X) is a C[h]-algebra.
Theorem (folklore)
Assume thatH2i+1(X) =0 for all i. Then:
(a) H2iC×+1(X) = 0 for alli.
Equivariant cohomology
If X is an algebraic variety endowed with a C×-action, one can define its equivariant cohomologyH•C×(X).
H•C×(pt)≃C[h], with deg(h) =2.
⇒ H•C×(X) is a C[h]-algebra.
Theorem (folklore)
Assume thatH2i+1(X) =0 for all i. Then:
(a) H2iC×+1(X) = 0 for alli.
(b) H2•(X)≃C[h]/hhi ⊗H2C•×(X).
Equivariant cohomology
If X is an algebraic variety endowed with a C×-action, one can define its equivariant cohomologyH•C×(X).
H•C×(pt)≃C[h], with deg(h) =2.
⇒ H•C×(X) is a C[h]-algebra.
Theorem (folklore)
Assume thatH2i+1(X) =0 for all i. Then:
(a) H2iC×+1(X) = 0 for alli.
(b) H2•(X)≃C[h]/hhi ⊗H2C•×(X).
(c) The canonical map iX∗ :H2•C×(X)−→H2•C×(XC×)is injective.
Equivariant cohomology
If X is an algebraic variety endowed with a C×-action, one can define its equivariant cohomologyH•C×(X).
H•C×(pt)≃C[h], with deg(h) =2.
⇒ H•C×(X) is a C[h]-algebra.
Theorem (folklore)
Assume thatH2i+1(X) =0 for all i. Then:
(a) H2iC×+1(X) = 0 for alli.
(b) H2•(X)≃C[h]/hhi ⊗H2C•×(X).
(c) The canonical map iX∗ :H2•C×(X)−→H2•C×(XC×)is injective.
Remark - H2C•×(XC×) =C[h]⊗H2•(X).
Equivariant cohomology of smooth Z
cEquivariant cohomology of smooth Z
cTheorem (B.-Shan 2017)
IfZc issmooth, then (a) H2iC×+1(Zc) =0.
(b) H2C•×(Zc)|{z}≃
C-alg
ReesFZ(CW).
Equivariant cohomology of smooth Z
cTheorem (B.-Shan 2017)
IfZc issmooth, then (a) H2iC×+1(Zc) =0.
(b) H2C•×(Zc)|{z}≃
C-alg
ReesFZ(CW).
Sketch of the proof. By folklore Theorem and Etingof-Ginzburg Theorem,
H2iC×+1(Zc) =0
Equivariant cohomology of smooth Z
cTheorem (B.-Shan 2017)
IfZc issmooth, then (a) H2iC×+1(Zc) =0.
(b) H2C•×(Zc)|{z}≃
C-alg
ReesFZ(CW).
Sketch of the proof. By folklore Theorem and Etingof-Ginzburg Theorem,
H2iC×+1(Zc) =0
H2•(Zc)≃
C[h]/hhi ⊗H2C•×(Zc)
Equivariant cohomology of smooth Z
cTheorem (B.-Shan 2017)
IfZc issmooth, then (a) H2iC×+1(Zc) =0.
(b) H2C•×(Zc)|{z}≃
C-alg
ReesFZ(CW).
Sketch of the proof. By folklore Theorem and Etingof-Ginzburg Theorem,
H2iC×+1(Zc) =0
H2•(Zc)≃
C[h]/hhi ⊗H2C•×(Zc) GradFZ(CW)
Equivariant cohomology of smooth Z
cTheorem (B.-Shan 2017)
IfZc issmooth, then (a) H2iC×+1(Zc) =0.
(b) H2C•×(Zc)|{z}≃
C-alg
ReesFZ(CW).
Sketch of the proof. By folklore Theorem and Etingof-Ginzburg Theorem,
H2iC×+1(Zc) =0
H2•(Zc)≃
C[h]/hhi ⊗H2C•×(Zc)
GradFZ(CW)≃C[h]/hhi ⊗ReesFZ(CW).
Sketch of the proof (continued)
H2•(Zc)≃
C[h]/hhi ⊗H2•C×(Zc)
GradFZ(CW)≃C[h]/hhi ⊗ReesFZ(CW).
Sketch of the proof (continued)
H2•(Zc)≃
C[h]/hhi ⊗H2•C×(Zc)
GradFZ(CW)≃C[h]/hhi ⊗ReesFZ(CW).
By folklore Theorem and Gordon Theorem, iZ∗c :H2C•×(Zc)֒−→H2C•×(ZC×
c )
Sketch of the proof (continued)
H2•(Zc)≃
C[h]/hhi ⊗H2•C×(Zc)
GradFZ(CW)≃C[h]/hhi ⊗ReesFZ(CW).
By folklore Theorem and Gordon Theorem, iZ∗c :H2C•×(Zc)֒−→H2C•×(ZC×
c )≃C[h]⊗Z(CW).
Sketch of the proof (continued)
H2•(Zc)≃
C[h]/hhi ⊗H2•C×(Zc)
GradFZ(CW)≃C[h]/hhi ⊗ReesFZ(CW).
By folklore Theorem and Gordon Theorem, iZ∗c :H2C•×(Zc)֒−→H2C•×(ZC×
c )≃C[h]⊗Z(CW).
An easy argument based on comparison of dimensions shows that it is sufficient to prove that
(⋆) ReesFZ(CW)⊂Im(iZ∗c).
Sketch of the proof (continued)
(⋆) ReesFZ(CW)⊂Im(iZ∗c).
Sketch of the proof (continued)
(⋆) ReesFZ(CW)⊂Im(iZ∗c).
IfE is a graded CW-module, then Hc⊗CW E is a graded projective Hc-module,
Sketch of the proof (continued)
(⋆) ReesFZ(CW)⊂Im(iZ∗c).
IfE is a graded CW-module, then Hc⊗CW E is a graded projective Hc-module, hence eHc⊗CW E is a graded projective Zc-module.
Sketch of the proof (continued)
(⋆) ReesFZ(CW)⊂Im(iZ∗c).
IfE is a graded CW-module, then Hc⊗CW E is a graded projective Hc-module, hence eHc⊗CW E is a graded projective Zc-module.
⇒equivariant Chern character ch(eHc⊗CW E)∈H^2C•×(Zc).
Sketch of the proof (continued)
(⋆) ReesFZ(CW)⊂Im(iZ∗c).
IfE is a graded CW-module, then Hc⊗CW E is a graded projective Hc-module, hence eHc⊗CW E is a graded projective Zc-module.
⇒equivariant Chern character ch(eHc⊗CW E)∈H^2C•×(Zc).
Theorem (Bellamy, 2009)
iZ∗c(ch(eHc ⊗CW E)) = X
χ∈Irr(W)
qbχhχ,C[V]co(W)⊗EigrW hχ,C[V]co(W)igrW eχW
where q =exp(h) and bχ=valhχ,C[V]co(W)igrW.
Sketch of the proof (continued)
(⋆) ReesFZ(CW)⊂Im(iZ∗c).
IfE is a graded CW-module, then Hc⊗CW E is a graded projective Hc-module, hence eHc⊗CW E is a graded projective Zc-module.
⇒equivariant Chern character ch(eHc⊗CW E)∈H^2C•×(Zc).
Theorem (Bellamy, 2009)
iZ∗c(ch(eHc ⊗CW E)) = X
χ∈Irr(W)
qbχhχ,C[V]co(W)⊗EigrW hχ,C[V]co(W)igrW eχW
where q =exp(h) and bχ=valhχ,C[V]co(W)igrW.
⇒(⋆) holds (!).
Sketch of the proof (continued)
(⋆) ReesFZ(CW)⊂Im(iZ∗c).
IfE is a graded CW-module, then Hc⊗CW E is a graded projective Hc-module, hence eHc⊗CW E is a graded projective Zc-module.
⇒equivariant Chern character ch(eHc⊗CW E)∈H^2C•×(Zc).
Theorem (Bellamy, 2009)
iZ∗c(ch(eHc ⊗CW E)) = X
χ∈Irr(W)
qbχhχ,C[V]co(W)⊗EigrW hχ,C[V]co(W)igrW eχW
where q =exp(h) and bχ=valhχ,C[V]co(W)igrW.
⇒(⋆) holds (!). QED
Back to example: W = S
nBack to example: W = S
nTheorem (B.-Maksimau, 2017)
Z1(Sn)µd =
Back to example: W = S
nTheorem (B.-Maksimau, 2017)
Z1(Sn)µd = a
d-coresγ s.t. d|n−|γ|
Zc
γ(G(d,1,rγ))
where rγ = (n−|γ|)/d.
Back to example: W = S
nTheorem (B.-Maksimau, 2017)
Z1(Sn)µd = a
d-coresγ s.t. d|n−|γ|
Zc
γ(G(d,1,rγ))
where rγ = (n−|γ|)/d. Fix such ad-coreγ.
Back to example: W = S
nTheorem (B.-Maksimau, 2017)
Z1(Sn)µd = a
d-coresγ s.t. d|n−|γ|
Zc
γ(G(d,1,rγ))
where rγ = (n−|γ|)/d.
Fix such ad-coreγ. By functoriality of equivariant cohomology, we get two maps
H2C•×(Z1(Sn)) ϕaγ
//
_
H2C•×(Zc
γ(G(d,1,rγ)))
_
H2C•×(Z1(Sn)C×) ϕbγ
//H2•C×(Zc
γ(G(d,1,rγ))C×)
Back to example: W = S
n(continued)
Back to example: W = S
n(continued)
This leads to a commutative diagram
ReesFZ(CSn) ϕaγ
//
_
ReesFZ(CG(d,1,rγ))
_
C[h]⊗Z(CSn) ϕbγ
//C[h]⊗Z(CG(d,1,rγ))
Back to example: W = S
n(continued)
This leads to a commutative diagram
ReesFZ(CSn) ϕaγ
//
_
ReesFZ(CG(d,1,rγ))
_
C[h]⊗Z(CSn) ϕbγ
//C[h]⊗Z(CG(d,1,rγ))
Using a theorem of Przezdziecki (2016), ϕbγ =IdC[h]⊗ϕγ (!).
Back to example: W = S
n(continued)
This leads to a commutative diagram
ReesFZ(CSn) ϕaγ
//
_
ReesFZ(CG(d,1,rγ))
_
C[h]⊗Z(CSn) ϕbγ
//C[h]⊗Z(CG(d,1,rγ))
Using a theorem of Przezdziecki (2016), ϕbγ =IdC[h]⊗ϕγ (!). QED
Some comments
Some comments
(1) One can deduce from this work a proof of a conjecture of
Ginzburg-Kaledin on the cohomology ofsymplectic resolutions of (V ×V∗)/W (B.-Shan, 2017).
Some comments
(1) One can deduce from this work a proof of a conjecture of
Ginzburg-Kaledin on the cohomology ofsymplectic resolutions of (V ×V∗)/W (B.-Shan, 2017).
(2) With R. Rouquier (2017), we have general conjectures about
◮ equivariant cohomology of possibly singular Calogero-Moser spaces,
Some comments
(1) One can deduce from this work a proof of a conjecture of
Ginzburg-Kaledin on the cohomology ofsymplectic resolutions of (V ×V∗)/W (B.-Shan, 2017).
(2) With R. Rouquier (2017), we have general conjectures about
◮ equivariant cohomology of possibly singular Calogero-Moser spaces,
◮ fixed points under the action of µd
Some comments
(1) One can deduce from this work a proof of a conjecture of
Ginzburg-Kaledin on the cohomology ofsymplectic resolutions of (V ×V∗)/W (B.-Shan, 2017).
(2) With R. Rouquier (2017), we have general conjectures about
◮ equivariant cohomology of possibly singular Calogero-Moser spaces,
◮ fixed points under the action of µd
+ The combinatoric fits with observations made in the representation theory of finite reductive groups
Some comments
(1) One can deduce from this work a proof of a conjecture of
Ginzburg-Kaledin on the cohomology ofsymplectic resolutions of (V ×V∗)/W (B.-Shan, 2017).
(2) With R. Rouquier (2017), we have general conjectures about
◮ equivariant cohomology of possibly singular Calogero-Moser spaces,
◮ fixed points under the action of µd
+ The combinatoric fits with observations made in the representation theory of finite reductive groups
◮ Lusztig families of unipotent characters
Some comments
(1) One can deduce from this work a proof of a conjecture of
Ginzburg-Kaledin on the cohomology ofsymplectic resolutions of (V ×V∗)/W (B.-Shan, 2017).
(2) With R. Rouquier (2017), we have general conjectures about
◮ equivariant cohomology of possibly singular Calogero-Moser spaces,
◮ fixed points under the action of µd
+ The combinatoric fits with observations made in the representation theory of finite reductive groups
◮ Lusztig families of unipotent characters
◮ partition into blocks of unipotent characters (d-Harish-Chandra theory)
Some comments
(1) One can deduce from this work a proof of a conjecture of
Ginzburg-Kaledin on the cohomology ofsymplectic resolutions of (V ×V∗)/W (B.-Shan, 2017).
(2) With R. Rouquier (2017), we have general conjectures about
◮ equivariant cohomology of possibly singular Calogero-Moser spaces,
◮ fixed points under the action of µd
+ The combinatoric fits with observations made in the representation theory of finite reductive groups
◮ Lusztig families of unipotent characters
◮ partition into blocks of unipotent characters (d-Harish-Chandra theory)
? Theory of spetses (Brou´e-Malle-Michel).