ope for conformal defects and holography
play

OPE for conformal defects and Holography Nozomu Kobayashi Kavli - PowerPoint PPT Presentation

OPE for conformal defects and Holography Nozomu Kobayashi Kavli IPMU, University of Tokyo Based on [1710.11165] with M. Fukuda, T. Nishioka and [1805.05967] with T. Nishioka (The Univ. of Tokyo) 1 Introduction Introduction Defects = Non-local


  1. OPE for conformal defects and Holography Nozomu Kobayashi Kavli IPMU, University of Tokyo Based on [1710.11165] with M. Fukuda, T. Nishioka and [1805.05967] with T. Nishioka (The Univ. of Tokyo) 1

  2. Introduction

  3. Introduction Defects = Non-local objects in QFTs • Defined by - boundary conditions around them T q - coupled to low-dimentional system � S 2 F = 2 πq • Many examples: 1-dim : Line operators (Wilson-’t Hooft loops) 2-dim : Surface operators D Codim-1 : Domain walls and boundaries T 1 T 2 Codim-2 : Entangling surface for entanglement entropy 2

  4. Why Defects? We can probe the part of theory which is inaccesible without defects - allow us to characterize the phase of theory • wilson loop in gauge theory • higher-form symmetry In fixed point ( CFT ), constrain bulk CFT data in defect-CFT by conformal bootstrap [Liendo-Rastelli-van Rees 12] O 1 O 2 O 1 O 2 � � = k O k l ˆ O l D D 3

  5. Conformal defects Especially, we consider particular class of defects: Conformal defects (codimension- m ) defects preserving SO ( d − m + 1 , 1) × SO ( m ) � �� � � �� � conformal sym. on defect rotational sym. around defect • conformal defects allow defect local operators ˆ O ( x ) • additional dynamical information appears - coupling to defects, defect local operators,... R d − m SO ( m ) 4 D ( m )

  6. OPE for conformal defects There are two types of OPE in Defect-CFT • Bulk-to-defect OPE : [Cardy 84, McAvity-Osborn 95] D ( m ) D ( m ) � n b ( m ) = + (descendants) ˆ O O n O ˆ O n • Defect OPE : [Berenstein-Corrado-Fischler-Maldacena 98, Gadde 16] D ( m ) � n c ( m ) + (descendants) = O n O n 5

  7. Several questions about DCFT 1. To what extent are we able to determine the structure of the defect OPE by conformal symmetry? - Decomposition by the irreducible representations � D ( m ) = B ( m ) [ O n ] n ∈ primaries 2. Can we probe the bulk AdS information by conformal defects on the boundary? H d +1 γ ( m ) R D ( m ) R d 3. Is there any extention of CFT? 6 - Spinning defects

  8. Overview of our results 1. Give the integral representation of the defect OPE blocks � d d x �O n ( x ) � D ( m ) ˜ B ( m ) [ O n ] = O n ( x ) O : shadow operator with ˜ ˜ ∆ = d − ∆ for O with ∆ 2. Reconstruct the AdS scalar field from the blocks • ˆ φ = B ( m ) [ O ] : B ( m ) [ O ] The Radon transform of the AdS scalar field φ when O scalar • Reproduce the (Euclidean) HKLL H d +1 φ p formula : � φ ( Y ) = φ (ˆ d d x K ( Y | x ) O ( x ) φ ) = 7

  9. Overview of our results 3. Study the kinematics and impliment of spinning defect in CFT: (1) calculating several correlators of bulk and defect local operators (2) exploring the OPE of spinning conformal defect (3) considering the correlators of two spinning conformal defects - deduced to thoese of scalar defects by recursion relation. 8

  10. Defect OPE blocks

  11. Defect OPE blocks We expect the defect OPE of the form �� � c ( m ) O n R ∆ n O n ( C ) + ( descendants ) D ( m ) ( P α ) = �D ( m ) ( P α ) � n ( R : radius , C : center vector , P α : vector to fix the defect ) D ( m ) � n c ( m ) + ( · · · ) = O n O n • The descendant terms are fixed by the primary O n and the conformal symmetry, � D ( m ) ( P α ) = B ( m ) [ P α , O n ] n • The defect OPE block is in the irreducible rep. of O n : �B ( m ) [ P α , O n ] O n ( X ) � = �D ( m ) ( P α ) O n ( X ) � . 9

  12. Projectors and shadows • Want to characterize the defect OPE blocks by their irreps • Spectral decomposition by the irreps of the conformal group: � 1 = |O n | n • |O n | : Projector onto the conformal multiplet of the primary O n [Ferrara-Grillo-Parisi-Gatto 72, · · · , Simmons-Duffin 12] For a scalar operator, � 1 D d X |O ∆ ( X ) � � ˜ |O ∆ | = O d − ∆ ( X ) | N ∆ ˜ O d − ∆ : the shadow operator of O ∆ Similary, the projector for spin l operator is investigated 10

  13. Integral representation of defect OPE blocks • Expand the defect by the projectors: �D ( m ) ( P α ) · · · � = �D ( m ) ( P α ) |O ∆ | · · · � + ( other irrep. ) � ∆ 1 � � D d X �D ( m ) ( P α ) O ∆ ( X ) � � ˜ = O d − ∆ ( X ) · · · � N ∆ ∆ + ( other irrep. ) • Can read off the block contribution: The integral rep. of the defect OPE block 1 � B ( m ) [ P α , O ∆ ] = D d X ˜ O d − ∆ ( X ) �D ( m ) ( P α ) O ∆ ( X ) � N ∆ 11

  14. Constraint equations There are two types of equations the defect OPE block satisfies The conformal Casimir equation � � L 2 ( P α ) + C ∆ ,l B ( m ) [ P α , O ∆ ,l ] = 0 • L 2 ( P α ) ≡ 1 2 L AB ( P α ) L AB ( P α ) : quadratic Casimir operator • C ∆ ,l = ∆(∆ − d ) + l ( l + d − 2) : the eigenvalue “Trivial” equations for scalar primaries O ∆ C ABCD ( P α ) B ( m ) [ P α , O ∆ ] = 0 2 L A [ B ( P α ) L CD ] ( P α ) : d ( d 2 − 1)( d +2) • C ABCD ( P α ) ≡ 1 quadratic 24 operators 12

  15. Moduli space of conformal defects • The moduli space has a coset structure: SO ( d + 1 , 1) M ( d,m ) = SO ( m ) × SO ( d + 1 − m, 1) • The quadratic Casimir operator is the Laplacian on M ( d,m ) − L 2 ( P α ) = � M ( d,m ) • The defect OPE block is a scalar field on M ( d,m ) Klein-Gordon equation on M ( d,m ) � � M ( d,m ) − M 2 � M 2 = C ∆ ,l B ( m ) [ P α , O ∆ ,l ] = 0 , 13

  16. Reconstruction of AdS scalar fields

  17. Conformal defects and submanifolds in AdS • Associated to a given defect D ( m ) is a unique submanifold γ ( m ) in AdS s.t. ∂γ ( m ) = D ( m ) • Their moduli spaces are equivalent: H d +1 γ ( m ) Isom ( AdS d +1 ) M ( d,m ) = Stab ( γ ( m ) ∈ AdS d +1 ) R D ( m ) R d • there is a map called Radon transform between Euclidean AdS ( = H d +1 ) and M ( d.m ) AdS d +1 M ( d,m ) ˆ ξ ˇ x ξ x 14

  18. Radon transform From H d +1 to M ( d,m ) : From M ( d,m ) to H d +1 : � � ˆ ˇ φ ( ξ ) = dν ( x ) φ ( x ) f ( x ) = dµ ( ξ ) f ( ξ ) ξ ∈ ˇ x ∈ ξ x • ξ : a codim- m submanifold in • ξ : a codim- m submanifold H d +1 through x • φ ( x ) : a function on H d +1 • f ( ξ ) : a function on M ( d,m ) Intertwining property � M ( d,m ) − M 2 � ˆ � � H d +1 − M 2 � � φ = 0 ⇔ φ = 0 15

  19. Reconstruct an AdS field from DOPE block • we can identify defect OPE block as the Radon transform of an AdS scalar field φ • Inversion formula for the Radon transform allows us to reconstruct φ from defect OPE block [Helgason 10] • Equivalent to the bulk reconstruction formula � D d X K ∆ ( Y | X ) O ∆ ( X ) φ ( Y ) = with the Euclidean version of the HKLL kernel [Hamilton-Kabat-Lifschytz-Lowe 06] M 2 = ∆( d − ∆) ( � H d +1 − M 2 ) K ∆ ( Y | X ) = 0 , 16

  20. Spinning conformal defect

  21. Spinning defects and recursion relation R d − m • Conformal defects can carry spin under SO ( m ) . • We adapt index-free notation for spinning defects introducing auxiliary transverse vector ˆ W , SO ( m ) D ( m ) W I 1 · · · ˆ ( ˆ W ) ≡ D ( m ) I 1 ··· I s ˆ W ◦ ˆ ˆ D ( m ) W I s , W = 0 s • In the same way as local operator case, we find recursion relation for one-point function, O ∆ ( X ) � = D s − s 0 ( ˆ �D ( m ) W ) �D ( m ) O ∆ ( X ) � s s 0 D s − s 0 ( ˆ W ) : s − s 0 -th order differential operator acting on P α 17

  22. Application of Recursion relation • Spinning defect OPE blocks � D ( m ) ( ˆ B ( m ) [ O n , ˆ W ) = W ] s s n ⇒ B ( m ) [ O ∆ ,l , ˆ W ] = D s − s 0 ( ˆ W ) B ( m ) s 0 [ O ∆ ,l ] s • Two-point function of spinnning defects �D ( m 1 ) ( ˆ W 1 ) D ( m 2 ) ( ˆ �B ( m 1 ) [ O n , ˆ W 1 ] B ( m 2 ) [ O n , ˆ � W 2 ) � = W 2 ] � s 1 s 2 s 1 s 2 n W 2 ) �D ( m 1 ) D ( m 2 ) �| spin- l ⇒ �D ( m 1 ) ( ˆ W 1 ) D ( m 2 ) ( ˆ W 2 ) �| spin- l = D s 1 ( ˆ W 1 ) D s 2 ( ˆ s 1 s 2 18

  23. Summary of results 1. Give the integral representation of the defect OPE blocks � d d x �O n ( x ) � D ( m ) ˜ B ( m ) [ O n ] = O n ( x ) O : shadow operator with ˜ ˜ ∆ = d − ∆ for O with ∆ 2. Reconstruct the AdS scalar field from the blocks • ˆ φ = B ( m ) [ O ] : B ( m ) [ O ] The Radon transform of the AdS scalar field φ when O scalar • Reproduce the (Euclidean) HKLL H d +1 φ p formula : � φ ( Y ) = φ (ˆ d d x K ( Y | x ) O ( x ) φ ) = 19

  24. Summary of results 3. Study the kinematics and impliment of spinning defect in CFT: (1) calculating several correlators of bulk and defect local operators (2) exploring the OPE of spinning conformal defect (3) considering the correlators of two spinning conformal defects - deduced to thoese of scalar defects by recursion relation. 20

  25. Future direction Integrable system The Casimir equation for the defect conformal block is shown to be equivalent to the Schrödinger equation of the Calogero-Sutherland model [Isachenkov-Liendo-Linke-Schomerus 18,...] - relation with our formalism? Holographic dual Can extend the construction to higher spin fields in AdS? - No known Radon transform beyond a scalar field in AdS - Spinning defects to incorporate spins 21

Recommend


More recommend