Trieste, Mar 24, 2016
Generalized Weyl anomalies in higher-spin theory
Wei Li (ITP-Chinese academy of sciences)
Generalized Weyl anomalies in higher-spin theory Wei Li - - PowerPoint PPT Presentation
Generalized Weyl anomalies in higher-spin theory Wei Li (ITP-Chinese academy of sciences) Trieste Mar 24, 2016 Main question CFT has conformal anomaly CFT with higher-spin symmetry has generalized conformal anomalies (in addition to
Trieste, Mar 24, 2016
Wei Li (ITP-Chinese academy of sciences)
✤ CFT has conformal anomaly ✤ CFT with higher-spin symmetry has generalized
✤ Today:
✤ Some aspects of holographic W-gravity
JHEP 1508, 035 (2015) with Stefan Theisen .
Even dimensional CFT in curved background gij, ϕijk, . . .
classical: T i
i = 0
W i
ij = 0
. . . quantum mechanical: hT i
ii 6= 0
hW i
iji 6= 0
. . .
Capper Duff ’73
Generating function of conformal anomalies
I Integrate out CFT fields to obtain (non-local) effective action
e−W [g] = Z DΦ e−SCFT[Φ,g]
I Weyl transformation
δσ2gij = 2 σ2 gij δσ2ϕijk = 4 σ2 ϕijk
δσ2W[g] = Z pg σ2(x) hT i
ii 6= 0
I Additional anomalous symmetry: W-Weyl transformation δσ3W[g, ϕ] = Z pg σ3(x) A3 6= 0
Even dimensional CFT in curved background gij, ϕijk, . . .
classical: T i
i = 0
W i
ij = 0
. . . quantum mechanical: hT i
ii 6= 0
hW i
iji 6= 0
. . .
Capper Duff ’73
Generating function of conformal anomalies
I Integrate out CFT fields to obtain (non-local) effective action
e−W [g,ϕ] = Z DΦ e−SCFT[Φ,g,ϕ]
I Weyl transformation
δσ2gij = 2 σ2 gij δσ2ϕijk = 4 σ2 ϕijk
δσ2W[g, ϕ] = Z pg σ2(x) A2 6= 0 I Additional anomalous symmetry: W-Weyl transformation δσ3W[g, ϕ] = Z pg σ3(x) A3 6= 0
Conserved currents
Tij(⌘ W (2)
ij )
W (3)
ijk
W (4)
ijkl
. . .
Naively
∂iTij = 0 and ηijTij = 0
Anomalous Ward Identity
hTij(p)Tkl(p)i = A(p2)
pkpl ηklp2
ηijTij = 0 = ) A(p2) = 0 = ) hTij(p)Tkl(p)i = 0
A(p2) = c p2
Conserved currents
Tij(⌘ W (2)
ij )
W (3)
ijk
W (4)
ijkl
. . .
Naively
∂iWi··· = 0 and ηijWij··· = 0
Anomalous Ward Identity
hWijk(p)Wlmn(p)i = A(3)(p2) ⇥ piplηilp2 pjpmηjmp2 pkpnηknp2 +. . . ⇤
ηijWijk = 0 = ) A(3)(p2) = 0 = ) hWijk(p)Wlmn(p)i = 0
A(3)(p2) = c(3) p2
OPE of holomorphic currents:
T(z)T(w) ∼ c (z − w)4 + . . . W(z)W(w) ∼ c(3) (z − w)6 + . . . . . . W (s)(z)W (s)(w) ∼ c(s) (z − w)2s + . . .
Each spin gives one Ws-Weyl anomaly
✤ Without higher-spin fields, 2D effective action is uniquely given
✤ Analogue of Polyakov action for other cases is not known
W2D[g] = Z R 1 ⇤R
Deser Schwimmer ’93; Deser ’96,’99
go on-shell
Weyl variation
Henningson Skenderis '98
go on-shell
Weyl variation
Henningson Skenderis '98
go on-shell
Weyl variation
Henningson Skenderis '98
go on-shell
Weyl variation
go on-shell
PBH transf. Imbimbo Schwimmer Theisen Yankielowicz '99
Step-1: Fefferman-Graham gauge of bulk metric (asympt. AdS2n+1)
ds2 = dρ2 ρ2 + 1 ρ2 gij(ρ, x)dxidxj FG expansion : gij(ρ, x) =
(0)
g ij(x) + ρ2(2) g ij(x) + ρ4(4) g ij(x) + . . .
Step-2: PBH transformation ≡ Bulk diffeo ξµ preserving FG gauge = Boundary Weyl Step-3: Weyl anomaly from PBH transformation on bulk action
hT i
i i =
⇣ δξSbulk⌘ |on-shell = 2 Z
∂M
ddx σ(x) h ρ p GL(G) i |ρ=0,on-shell
Step-4: rewrite in terms of boundary data (i.e. go on-shell)
go on-shell
PBH transf. Imbimbo Schwimmer Theisen Yankielowicz '99
✤ gravity theory coupled to higher-spin gauge
✤ dual CFT with higher-spin currents
✤ Higher-spin gravity (Vasiliev’s theory) is an
✤ Holography with higher-spin symmetry is different
✤ Can use higher-spin symmetry to study string theory
Campoleoni Fredenhagen Pfenninger Theisen '12
Action:
S = SCS[A] − SCS[ ˜ A] with SCS[A] = k 4π Z
M
Tr[AdA + 2 3A3] Lorentzian: A, ˜ A ∈ sl(N, R) Euclidean: A, ˜ A ∈ sl(N, C) and ˜ A = −A†
Translation to metric-like formalism
e = A − ˜ A 2 ω = A + ˜ A 2
Gµν = Tr[eµeν] ϕµνρ = Tr[e{µeνeρ}] . . .
Spectrum
spin-2 : {L1, L0, L−1}
spin-s : {W (s)
m }
m = −s + 1, . . . , s − 1
Principal embedding: 1 spin-s field for each s = 2, ..., N
L1 L0 L−1 W2 W1 W0 W−1 W−2 highest/lowest weight modes
Spectrum
spin-2 : {L1, L0, L−1}
spin-s : {W (s)
m }
m = −s + 1, . . . , s − 1
Principal embedding: 1 spin-s field for each s = 2, ..., N
L1 L0 L−1 W2 W1 W0 W−1 W−2 lowest weight/zero/highest weight modes
not at the level of action
Li Theisen '15
Li Theisen '15
Step-1: Fefferman-Graham gauge of sl(2) (i.e. pure gravity)
A(ρ, x) = ρL0 a(x) ρ−L0 − dρ
ρ L0
˜ A(ρ, x) = ρ−L0 ˜ a(x) ρL0 + dρ
ρ L0
with Tr[L0(a − ˜ a)] = 0
Step-2: PBH transformation for sl(2):
U(ρ, x) = ρL0 u(x) ρ−L0 with u2 = σ2L0
Step-3: Weyl anomalies from PBH transformation on bulk action
Weyl Anomaly = c Tr ⇥ L0
z − ¯
∂az ⇤
Step-4: rewrite in terms of boundary data
Step-1: Fefferman-Graham gauge of sl(N)
A(ρ, x) = ρL0 a(x) ρ−L0 − dρ
ρ L0
˜ A(ρ, x) = ρ−L0 ˜ a(x) ρL0 + dρ
ρ L0
with Tr[L0(a − ˜ a)] = 0 and Tr[W0(a − ˜ a)] = 0 . . .
Step-2: PBH transformation for sl(N):
U(ρ, x) = ρL0 u(x) ρ−L0 with u2 = σ2L0 and u3 = σ3W0 . . .
Step-3: Weyl anomalies from PBH transformation on bulk action
Weyl Anomaly = c Tr ⇥ L0
z − ¯
∂az ⇤ W3-Weyl Anomaly = c Tr ⇥ W0
z − ¯
∂az ⇤ . . .
Step-4: rewrite in terms of boundary data
Li Theisen '15
Conformal gauge
I No source turned on; N − 1 conformal mode: Toda fields {Φs} I Effective action is local (Toda action) I Boundary metric and spin-3 field (ΨL ≡ 1
2 (Φ1 + Φ2)
ΨW ≡ 1
2 (Φ1 − Φ2))
g = eΨL cosh(ΨW ) dzd¯ z and ϕ = 0
Weyl anomaly A2 = c 6√g ∂ ¯ ∂ΨL W-Weyl anomaly A3 = c 18√g ∂ ¯ ∂ΨW
Light-cone gauge
I Turn on (chiral) sources µ2, µ3 (¯ µs = 0) Tzz = δ δµ2 and Wzzz = δ δµ3 I Boundary metric and spin-3 field g = (dz + µ2d¯ z) d¯ z and ϕ = µ3 d¯ z3
Weyl anomaly A2 = c 3 ∂2µ2 W-Weyl anomaly A3 = c 18 ∂(∂2 − T)µ3
− c
12R[g]
c Tr ⇥ L0
z − ¯
∂az ⇤ c Tr ⇥ L0
z − ¯
∂az ⇤ c Tr ⇥ W0
z − ¯
∂az ⇤ conformal
c 6√g ∂ ¯
∂ΨL light-cone
c 3 ∂2µ2
conformal
c 18√g ∂ ¯
∂ΨW light-cone
c 18 ∂(∂2 − T)µ3
Weyl anomaly W-Weyl anomaly pure gravity metric pure gravity sl(2) higher-spin sl(N) higher-spin metric-like higher-spin metric-like covariant ? ?
✤ Weyl anomaly and W-Weyl anomaly ✤ adapt PBH procedure to sl(N) Chern-Simons theory ✤ conformal gauge and light-cone gauge
✤ translate anomalies into (covariant expression of)
✤ effective action in terms of metric and higher-spin
✤ 4d?