Loop corrections to supergravity Agnese Bissi Harvard University - - PowerPoint PPT Presentation

loop corrections to supergravity
SMART_READER_LITE
LIVE PREVIEW

Loop corrections to supergravity Agnese Bissi Harvard University - - PowerPoint PPT Presentation

Loop corrections to supergravity Agnese Bissi Harvard University based on: 1706.02388 w/Alday 1612.03891 w/Aharony, Alday and Perlmutter AdS/CFT correspondence: Gravitational theory in CFT in D dimensions D+1 dimensions G N N Newton


slide-1
SLIDE 1

Loop corrections to supergravity

Agnese Bissi

Harvard University based on: 1706.02388 w/Alday 1612.03891 w/Aharony, Alday and Perlmutter

slide-2
SLIDE 2

2

AdS/CFT correspondence: CFT in D dimensions Gravitational theory in D+1 dimensions GN N

Newton Constant rank of the gauge group

Loops in AdS expansion 1 N 2

Use structure and symmetries (superconformal symmetry)

  • f the CFT to compute loop amplitudes in AdS
slide-3
SLIDE 3

+ + + + +… G =

{ {

{

3

General idea: N = 4 SYM with SU(N) gauge group at large N 0 1 N 2 1 N 4

λ = g2N

Generalised free fields Supergravity Supergravity loops

slide-4
SLIDE 4

+ + + + +… G =

{ {

{

3

General idea: N = 4 SYM with SU(N) gauge group at large N 0 1 N 2 1 N 4

λ = g2N

Generalised free fields Supergravity Supergravity loops

slide-5
SLIDE 5

4

  • Consider the superconformal primary with protected

dimension , transforming in the of R- symmetry BPS operator ∆ = 2 [0, 2, 0] SU(4)

1 2

O2 hO2(x1)O2(x2)O2(x3)O2(x4)i = X

R

GR(u, v) x4

12x4 34

slide-6
SLIDE 6

4

  • Consider the superconformal primary with protected

dimension , transforming in the of R- symmetry BPS operator ∆ = 2 [0, 2, 0] SU(4)

1 2

all the reps in

[0, 2, 0] × [0, 2, 0]

cross ratios O2 hO2(x1)O2(x2)O2(x3)O2(x4)i = X

R

GR(u, v) x4

12x4 34

slide-7
SLIDE 7

4

  • Consider the superconformal primary with protected

dimension , transforming in the of R- symmetry BPS operator ∆ = 2 [0, 2, 0] SU(4)

1 2

O2 hO2(x1)O2(x2)O2(x3)O2(x4)i = X

R

GR(u, v) x4

12x4 34

slide-8
SLIDE 8

4

  • Consider the superconformal primary with protected

dimension , transforming in the of R- symmetry BPS operator ∆ = 2 [0, 2, 0] SU(4)

1 2

  • Superconformal Ward identities imply relations among

allowing the entire four point function to be expressed in terms of a non trivial function GR G(u, v) O2 hO2(x1)O2(x2)O2(x3)O2(x4)i = X

R

GR(u, v) x4

12x4 34

slide-9
SLIDE 9

OPE content O2 × O2 ∼ Olong + Oshort

Oshort

Olong

∆L = ∆(gY M, N) cL = cO2O2Olong(gY M, N) cS = cO2O2Oshort(N) ∆S = ∆(N) acquire anomalous dimension protected ops (1/2 and 1/4 BPS) X

OL OS

cL cL

cS cS

G(u, v) =

slide-10
SLIDE 10

hO20(x1)O20(x2)O20(x3)O20(x4)i = hO20(x1)O20(x2)O20(x3)O20(x4)i

6

v2G(u, v) − u2G(v, u) + 4(u2 − v2) + 16(u − v) N 2 − 1 = 0 = X

OL OS

X

OL OS

cS

cL cL

cS

cL cL

cS cS

slide-11
SLIDE 11

7

G(u, v) = Gshort(u, v) + H(u, v) Since are protected, this function is completely determined! OS X

OL

c2

LgOL

= X

OL OS

X

OL OS

cS

cL cL

cS

cL cL

cS cS

The sum runs over superconformal primaries, which are singlet of and are superconformal blocks.

SU(4)R

gOL

slide-12
SLIDE 12

Properties of conformal blocks

  • Conformal blocks are fixed by conformal symmetry and

encode the contribution of a primary and all its descendants. In 4d they are known in a closed form, for scalar external

  • perators.
  • They are eigenfunctions of the quadratic Casimir operator

with eigenvalue ∆, `

  • Note that ( ) label intermediate operators and denote the

conformal dimension and spin, respectively.

J2 ∼ `2

8

CG∆,`(u, v) = J2G∆,`(u, v)

slide-13
SLIDE 13

Properties of conformal blocks II

9

  • Small u limit

⌧ = ∆ − ` where

  • Small v limit
  • Superconformal blocks

G∆,`(u, v) → u

τ 2 ˜

gcoll

∆,`(v) + . . .

G∆,`(u, v) = u

τ 2 ˜

g∆,`(u, v) ˜ g∆,`(u, v) → log(v) g∆,`(u, v) = u

τ 2 ˜

g∆+4,`(u, v)

slide-14
SLIDE 14

10

Large N expansion H(u, v) = H(0)(u, v) + 1 N 2 H(1)(u, v) + 1 N 4 H(2)(u, v) c2

∆,` = a∆,` = a(0) ∆,` + 1

N 2 a(1)

∆,` + 1

N 4 a(2)

∆,`(u, v)

∆ = ∆(0) + 1 N 2 γ(1) + 1 N 4 γ(2)

slide-15
SLIDE 15

10

Large N expansion H(u, v) = H(0)(u, v) + 1 N 2 H(1)(u, v) + 1 N 4 H(2)(u, v) c2

∆,` = a∆,` = a(0) ∆,` + 1

N 2 a(1)

∆,` + 1

N 4 a(2)

∆,`(u, v)

∆ = ∆(0) + 1 N 2 γ(1) + 1 N 4 γ(2)

slide-16
SLIDE 16

11

Order For large and large the result reduces to the one of generalised free fields: λ N G(0)(u, v) = 4 ✓ 1 + 1 v2 ◆ From the OPE content of protected operators G(0)(u, v) = G(0)

short(u, v) +

X

n,`

a(0)

n,`un+2gn,`(u, v)

∆(0) = 2n + ` + 4 a(0)

n,`

N 0 H(0)(u, v) Double trace operators! O2⇤n∂`O2

slide-17
SLIDE 17

Comments on large spin limit

12

For small u and v: LHS RHS H(0)(u, v) = X

∆,`

a∆,`u

∆−` 2 g∆,`(u, v)

slide-18
SLIDE 18

Comments on large spin limit

12

H(0)(u, v) → 2 3 u2 v2 + . . . For small u and v: LHS RHS H(0)(u, v) = X

∆,`

a∆,`u

∆−` 2 g∆,`(u, v)

slide-19
SLIDE 19

Comments on large spin limit

12

H(0)(u, v) → 2 3 u2 v2 + . . . For small u and v: LHS RHS Power law divergence! H(0)(u, v) = X

∆,`

a∆,`u

∆−` 2 g∆,`(u, v)

slide-20
SLIDE 20

Comments on large spin limit

12

H(0)(u, v) → 2 3 u2 v2 + . . . For small u and v: LHS RHS Power law divergence!

  • Each conformal block

diverges as log(v)

  • Need of an infinite

sum on the spin, of

  • perators whose twist

approaches . H(0)(u, v) = X

∆,`

a∆,`u

∆−` 2 g∆,`(u, v)

∆ − ` = 4

slide-21
SLIDE 21

Comments on large spin limit

12

H(0)(u, v) → 2 3 u2 v2 + . . . For small u and v: LHS RHS Power law divergence!

  • Each conformal block

diverges as log(v)

  • Need of an infinite

sum on the spin, of

  • perators whose twist

approaches . Fix dimensions and OPE coefficients to all

  • rders in

1/` H(0)(u, v) = X

∆,`

a∆,`u

∆−` 2 g∆,`(u, v)

∆ − ` = 4

slide-22
SLIDE 22

13

Large N expansion H(u, v) = H(0)(u, v) + 1 N 2 H(1)(u, v) + 1 N 4 H(2)(u, v) ∆ = ∆(0) + 1 N 2 γ(1) + 1 N 4 γ(2) a∆,` = a(0)

∆,` + 1

N 2 a(1)

∆,` + 1

N 4 a(2)

∆,`

slide-23
SLIDE 23

13

Large N expansion H(u, v) = H(0)(u, v) + 1 N 2 H(1)(u, v) + 1 N 4 H(2)(u, v) ∆ = ∆(0) + 1 N 2 γ(1) + 1 N 4 γ(2) a∆,` = a(0)

∆,` + 1

N 2 a(1)

∆,` + 1

N 4 a(2)

∆,`

slide-24
SLIDE 24

14

Order N −2 + Supergravity result: G(1)(u, v) = −16u2 ¯ D2422(u, v) = G(1)

short(u, v) + H(1)(u, v)

Absence of new operators only corrections to the dimensions and the OPE coefficients of double trace operators γ(1)

n,`

a(1)

n,` = 1

2 ∂ ∂n ⇣ a(0)

n,`γ(1) n,`

H(1)(u, v) = X

n,`

u2+n ✓ a(1)

n,` + 1

2a(0)

n,`γ(1) n,`(log u + ∂

∂n) ◆ g4+2n+`,`(u, v)

slide-25
SLIDE 25

Alternative method Crossing equation: v2G(u, v) − u2G(v, u) + 4(u2 − v2) + 16(u − v) N 2 − 1 = 0

  • protected operators
  • no

15

log u

  • double trace operators

n ≥ 0 G(1)(u, v) = G(1)

short(u, v) + H(1)(u, v)

τ = 4 + 2n

slide-26
SLIDE 26

Equation for H(1)(u, v)

16

H(1)(u, v) = u2 v2 H(1)(v, u) − 4(u2 − v2) v2 − 16 v2(N 2 − 1) +u2 v2 G(1)

short(v, u) − G(1) short(u, v)

slide-27
SLIDE 27

Equation for H(1)(u, v)

16

H(1)(u, v) = u2 v2 H(1)(v, u) − 4(u2 − v2) v2 − 16 v2(N 2 − 1) +u2 v2 G(1)

short(v, u) − G(1) short(u, v)

slide-28
SLIDE 28

Equation for H(1)(u, v)

16

H(1)(u, v) = u2 v2 H(1)(v, u) − 4(u2 − v2) v2 − 16 v2(N 2 − 1) +u2 v2 G(1)

short(v, u) − G(1) short(u, v)

u2 v2 H(1)(v, u)

  • no divergence!
  • all the divergences in v come from the protected part

= u2 v2 X

n,`

a(0)

n,`v2+nγ(1) n,`g4+2n+`,`(v, u) log v + . . .

slide-29
SLIDE 29

Equation for H(1)(u, v)

16

H(1)(u, v) = u2 v2 H(1)(v, u) − 4(u2 − v2) v2 − 16 v2(N 2 − 1) +u2 v2 G(1)

short(v, u) − G(1) short(u, v)

H(u, v)|log u = X

n,`

a(0)

n,`u2+nγ(1) n,`g4+2n+`,`(u, v)

= u2 v2 G(1)

short(v, u)|log u

slide-30
SLIDE 30

Equation for H(1)(u, v)

16

H(1)(u, v) = u2 v2 H(1)(v, u) − 4(u2 − v2) v2 − 16 v2(N 2 − 1) +u2 v2 G(1)

short(v, u) − G(1) short(u, v)

H(u, v)|log u = X

n,`

a(0)

n,`u2+nγ(1) n,`g4+2n+`,`(u, v)

= u2 v2 G(1)

short(v, u)|log u

  • match the divergences on both sides to all orders in

γ(1)

n,`

1/`

slide-31
SLIDE 31

17

Large N expansion H(u, v) = H(0)(u, v) + 1 N 2 H(1)(u, v) + 1 N 4 H(2)(u, v) ∆ = ∆(0) + 1 N 2 γ(1) + 1 N 4 γ(2) a∆,` = a(0)

∆,` + 1

N 2 a(1)

∆,` + 1

N 4 a(2)

∆,`

slide-32
SLIDE 32

17

Large N expansion H(u, v) = H(0)(u, v) + 1 N 2 H(1)(u, v) + 1 N 4 H(2)(u, v) ∆ = ∆(0) + 1 N 2 γ(1) + 1 N 4 γ(2) a∆,` = a(0)

∆,` + 1

N 2 a(1)

∆,` + 1

N 4 a(2)

∆,`

slide-33
SLIDE 33

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

Order N −4 Idea: extract information about using γ(2)

n,`

  • CFT data from previous orders
  • crossing symmetry

General picture:

18

slide-34
SLIDE 34

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

Order N −4 General picture:

18

slide-35
SLIDE 35

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

Order N −4 log2 u with coefficient fixed by known CFT data General picture:

18

slide-36
SLIDE 36

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

Order N −4 log2 u with coefficient fixed by known CFT data General picture:

18

slide-37
SLIDE 37

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

Order N −4 log2 u with coefficient fixed by known CFT data General picture:

18

CROSSING
 SYMMETRY

slide-38
SLIDE 38

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

Order N −4 log2 u with coefficient fixed by known CFT data General picture:

18

CROSSING
 SYMMETRY log2 v extract to all orders in γ(2)

n,`

1/`

slide-39
SLIDE 39

It is not so “simple”…

19

More than one operator with the same , and R-symmetry ` τ (1) H2(u, v)|log2 u = 1 8 X

n,`,I

un+2a(0)

n,`,I(γ(1) n,`,I)2g4+n+`,`

does NOT follow from Solve the mixing problem!! {O2⇤n∂`O2, O3⇤n−1∂`O3, O4⇤n−2∂`O4, . . . } X a(0)

n,`

X a(0)

n,`γ(1) n,`

slide-40
SLIDE 40

20

Mixing problem {O2⇤n∂`O2, O3⇤n−1∂`O3, O4⇤n−2∂`O4, . . . } intermediate operators in the four point functions Data (Dolan Osborn, Uruchurtu, Rastelli Zhou): hOpOpOqOqi X

I

c(0)

ppIc(0) qqI

X

I

c(0)

ppIc(0) qqIγ(1) I

X

I

c(0)

ppIc(0) qqI(γ(1) I )2

hOpOpOqOqi(0) hOpOpOqOqi(1)

slide-41
SLIDE 41

Example

21

{O2⇤2∂`O2, O3⇤∂`O3, O4∂`O4} ∆(0) = 8 + ` hO2O2O2O2i hO2O2O3O3i hO3O3O3O3i hO2O2O4O4i hO3O3O4O4i hO4O4O4O4i n = 2 6 x 2 X

I

c(0)

ppIc(0) qqI

X

I

c(0)

ppIc(0) qqIγ(1) I

slide-42
SLIDE 42

Example

21

{O2⇤2∂`O2, O3⇤∂`O3, O4∂`O4} ∆(0) = 8 + ` hO2O2O2O2i hO2O2O3O3i hO3O3O3O3i hO2O2O4O4i hO3O3O4O4i hO4O4O4O4i n = 2 6 x 2 X

I

c(0)

ppIc(0) qqI

X

I

c(0)

ppIc(0) qqIγ(1) I

3+3+3 N 0 N −2 and 3

slide-43
SLIDE 43

Example

21

{O2⇤2∂`O2, O3⇤∂`O3, O4∂`O4} ∆(0) = 8 + `

  • # of operators: 3
  • # of 4 point functions =eqs : 12
  • # of unknown: 12

hO2O2O2O2i hO2O2O3O3i hO3O3O3O3i hO2O2O4O4i hO3O3O4O4i hO4O4O4O4i n = 2 6 x 2 X

I

c(0)

ppIc(0) qqI

X

I

c(0)

ppIc(0) qqIγ(1) I

3+3+3 N 0 N −2 and MIXING PROBLEM CAN BE SOLVED! 3

slide-44
SLIDE 44

22

Solving the mixing problem

  • 1. From the explicit results we get the structure:

2.Looking at the term fix all the coefficients! (n + 1)3(n + 2)3(n + 3)3(n + 4)3(5 + 2n) 120(J2 − (n + 2)(n + 3))2 +

n+2

X

j=2

βn,j J2 − j(j + 1) P

I a(0) n,`,I(γ(1) n,`,I)2

P

I a(0) n,`,I

log2 u log2 v H(2)(u, v)

  • log2 u log2 v

crossing symmetric! Fix the average to all orders in 1/`

slide-45
SLIDE 45

22

Solving the mixing problem

  • 1. From the explicit results we get the structure:

2.Looking at the term fix all the coefficients! (n + 1)3(n + 2)3(n + 3)3(n + 4)3(5 + 2n) 120(J2 − (n + 2)(n + 3))2 +

n+2

X

j=2

βn,j J2 − j(j + 1) P

I a(0) n,`,I(γ(1) n,`,I)2

P

I a(0) n,`,I

log2 u log2 v H(2)(u, v)

  • log2 u log2 v

crossing symmetric! Fix the average to all orders in 1/`

slide-46
SLIDE 46

23

Nicer way

h(γ(1)

n,`)2i = ∞

X

p=2 p−1

Y

k=2

αp(n + 1)2(n + 2)2(n + 3)2(n + 4)2(n k + 2)(n + k + 3) (J2 (n + 2)(n + 3))2(J2 k(k + 1))

Each term is the contribution of a Kaluza Klein mode! Op⇤n∂`Op Next step: contribution of each KK mode to γ(2)

0,`

slide-47
SLIDE 47

24

Full expansion

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

  • Small u, v limit
  • Focus on the terms proportional to

log u log2 v

X

`

1 4a(0)

0,`(γ(1) 0,` )2∂ngn,`(u, v)

  • n=0

+ 1 2 ✓ a(0)

0,`γ(2) 0,` + 1

2a(1)

0,`γ(1) 0,`

◆ gcoll

0,` (v)

  • log2 v

= X

n,`

1 8a(0)

n,`(γ(1) n,`)2vngn,`(v, u)

  • u0 log u
slide-48
SLIDE 48

24

Full expansion

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

  • Small u, v limit
  • Focus on the terms proportional to

log u log2 v

X

`

1 4a(0)

0,`(γ(1) 0,` )2∂ngn,`(u, v)

  • n=0

+ 1 2 ✓ a(0)

0,`γ(2) 0,` + 1

2a(1)

0,`γ(1) 0,`

◆ gcoll

0,` (v)

  • log2 v

= X

n,`

1 8a(0)

n,`(γ(1) n,`)2vngn,`(v, u)

  • u0 log u
slide-49
SLIDE 49

24

Full expansion

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

  • Small u, v limit
  • Focus on the terms proportional to

log u log2 v

X

`

1 4a(0)

0,`(γ(1) 0,` )2∂ngn,`(u, v)

  • n=0

+ 1 2 ✓ a(0)

0,`γ(2) 0,` + 1

2a(1)

0,`γ(1) 0,`

◆ gcoll

0,` (v)

  • log2 v

= X

n,`

1 8a(0)

n,`(γ(1) n,`)2vngn,`(v, u)

  • u0 log u
slide-50
SLIDE 50

24

Full expansion

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

  • Small u, v limit
  • Focus on the terms proportional to

log u log2 v (u, v) → (v, u)

X

`

1 4a(0)

0,`(γ(1) 0,` )2∂ngn,`(u, v)

  • n=0

+ 1 2 ✓ a(0)

0,`γ(2) 0,` + 1

2a(1)

0,`γ(1) 0,`

◆ gcoll

0,` (v)

  • log2 v

= X

n,`

1 8a(0)

n,`(γ(1) n,`)2vngn,`(v, u)

  • u0 log u
slide-51
SLIDE 51

24

Full expansion

H(2)(u, v) = X

n,`

a(2)

n,` + 1

2a(0)

n,`γ(2) n,`∂n + 1

2a(1)

n,`γ(1) n,`∂n

+ 1 8a(0)

n,`(γ(1) n,`)2∂2 n

! u2+ngn,`(u, v)

  • Small u, v limit
  • Focus on the terms proportional to

log u log2 v

X

`

1 4a(0)

0,`(γ(1) 0,` )2∂ngn,`(u, v)

  • n=0

+ 1 2 ✓ a(0)

0,`γ(2) 0,` + 1

2a(1)

0,`γ(1) 0,`

◆ gcoll

0,` (v)

  • log2 v

= X

n,`

1 8a(0)

n,`(γ(1) n,`)2vngn,`(v, u)

  • u0 log u
slide-52
SLIDE 52

X

`

1 4a(0)

0,`(γ(1) 0,` )2∂ngn,`(u, v)

  • n=0

+ 1 2 ✓ a(0)

0,`γ(2) 0,` + 1

2a(1)

0,`γ(1) 0,`

◆ gcoll

0,` (v)

  • log2 v

= X

n,`

1 8a(0)

n,`(γ(1) n,`)2vngn,`(v, u)

  • u0 log u

Technical complications:

  • deal with derivatives of the blocks
  • infinite sum with insertion of (γ(1)

n,`)2

25

Next: Plug the expression for in terms of KK modes and see the contribution of each KK mode to (γ(1)

n,`)2

γ(2)

0,`

slide-53
SLIDE 53

26

γ(2)

0,` − 1

2γ(1)

0,` ∂nγ(1) n,`

  • p

=αp P (14+2`)(p) (p2 − 4)(p2 − 1)p + αp(p2 − 4)(p2 − 1)p3Q(4+2`)(p)ψ(2)(p)

polynomials ∆0,2 =6 − 4 N 2 − 45 N 4 + · · · ∆0,4 =8 − 48 25 1 N 2 − 12768 3125 1 N 4 + · · · ∼ p1−2` for large p sum over p diverges for spin 0!

slide-54
SLIDE 54

26

γ(2)

0,` − 1

2γ(1)

0,` ∂nγ(1) n,`

  • p

=αp P (14+2`)(p) (p2 − 4)(p2 − 1)p + αp(p2 − 4)(p2 − 1)p3Q(4+2`)(p)ψ(2)(p)

polynomials ∆0,2 =6 − 4 N 2 − 45 N 4 + · · · ∆0,4 =8 − 48 25 1 N 2 − 12768 3125 1 N 4 + · · · ∼ p1−2` for large p sum over p diverges for spin 0! first loop correction to the 4 graviton superamplitude!

slide-55
SLIDE 55

27

Comments and outlook

  • Divergence at spin 0:

quadratic divergence in 10d sugra, at large with only spin zero support λ

  • Going beyond n=0
  • Extend to other 4 point functions, e.g. external operator with
  • Extend to other susy theory

∆ = 3