Boundary flow and geometric realization in holographic $T\bar T$-deformed BCFT

Technical Summary: Boundary Flow and Geometric Realization in Holographic $T\bar{T}$-Deformed BCFT

Problem Statement
The paper addresses the conceptual and technical subtleties arising when applying the $T\bar{T}$ deformation to Boundary Conformal Field Theories (BCFTs). While the $T\bar{T}$ deformation of two-dimensional CFTs is well-understood via holography (often modeled by a finite radial cutoff in AdS$_3$), the presence of a physical boundary introduces complications. Specifically, there is ambiguity regarding how the deformation interacts with the boundary: does it induce a physical displacement of the boundary, introduce new boundary degrees of freedom, or merely reorganize existing data? Furthermore, different holographic realizations (cutoff geometries) appear to yield different physical pictures (e.g., moving vs. fixed boundaries), necessitating a rigorous distinction between intrinsic field-theoretic definitions and specific geometric implementations.

Methodology
The authors employ a dual approach, combining an intrinsic field-theoretic formulation with holographic analysis in AdS$_3$/BCFT$_2$:

1. Intrinsic Field-Theoretic Formulation:

   • The deformation is defined not by adding a composite operator to the action, but by modifying the asymptotic variational principle of AdS$_3$ gravity. This involves imposing a mixed asymptotic boundary condition that fixes a specific nonlinear combination of the boundary metric and its subleading terms, rather than the metric alone.
   • This definition yields an exact, quadratic trace relation for the stress tensor without introducing a finite radial cutoff.
   • The analysis restricts attention to BCFTs without independent boundary degrees of freedom (standard Cardy-type BCFTs), where all boundary responses are fixed by Ward identities.
   • The authors utilize boundary Ward identities and the displacement operator ($D$), which is the normal-normal component of the stress tensor at the boundary ($T_{nn}|_\Sigma$), to derive the flow equations.

2. Holographic Realizations:

   • Two distinct geometric realizations of the same intrinsic deformation are analyzed:
     • Type A: A rigid Dirichlet cutoff surface intersects the End-of-the-World (EOW) brane at a finite position.
     • Type B: The cutoff surface is chosen such that its induced metric is asymptotically AdS$_2$, causing it to intersect the EOW brane only at asymptotic infinity.
   • Entanglement Entropy is computed for both types at zero and finite temperature (BTZ background) to serve as a quantitative diagnostic, comparing holographic results with field-theoretic expectations.

Key Contributions and Results

1. Boundary-Localized Flow:

   • The authors demonstrate that for BCFTs without independent boundary degrees of freedom, the bulk $T\bar{T}$ composite operator collapses to a purely boundary-localized flow.
   • Under reflective boundary conditions ($T_{tx}|_\Sigma = 0$), the deformation is governed entirely by the displacement operator $D$. The bulk operator reduces to a universal one-dimensional functional supported on the boundary.
   • An exact flow equation for the generating functional $W_\lambda$ is derived:
     $$\left(\frac{\partial W_\lambda}{\partial \lambda}\right)_\Sigma = -\int_\Sigma dt \, \ell_\lambda \frac{D^2}{1 - \pi \lambda D}$$
     where $\ell_\lambda \propto \sqrt{\lambda}$ is the intrinsic length scale. This flow can be integrated in closed form to yield an induced boundary action, showing that the deformation reorganizes existing boundary data without introducing new degrees of freedom.

2. Fixed–Moving Boundary Equivalence:

   • A precise equivalence is established between two descriptions of the deformation:
     • Fixed-Boundary: The boundary embedding is fixed, and the deformation is encoded in a nonlinear boundary action (induced by the flow).
     • Moving-Boundary: The boundary condition is fixed, but the physical boundary embedding undergoes a $\lambda$-dependent shift.
   • These are shown to be related by a Legendre-type transformation (variational transformation). They represent different parameterizations of the same variational problem, not distinct physical dynamics. The displacement operator acts as the conjugate variable mediating this equivalence.

3. Holographic Realizations (Type A vs. Type B):

   • Type A (Boundary Deformed): The rigid cutoff intersects the EOW brane at a finite, $\lambda$-dependent position. This results in an apparent finite displacement of the physical BCFT boundary. Here, the displacement operator $D$ acquires a non-zero expectation value.
   • Type B (Boundary Preserved): The cutoff surface is asymptotically AdS$_2$. Consequently, the physical BCFT boundary coincides with the conformal boundary of AdS$_2$ and is geometrically pinned. No finite boundary displacement occurs, and the displacement operator vanishes identically ($D=0$).
   • Despite these geometric differences, the authors show that entanglement entropy calculations in both Type A and Type B yield identical results (matching BCFT expectations). This confirms that the universal consequences of the intrinsic flow are independent of the specific holographic implementation.

4. Scaling and Divergence:

   • The paper clarifies the scaling behavior of the displacement operator. In the static holographic saddle (Type A), as the deformation parameter $\lambda \to 0$, the boundary shift $\delta x$ vanishes as $\sqrt{\lambda}$, but the conjugate displacement density $D$ diverges as $1/\lambda$. This indicates that the holographic saddle approaches a "rigid-boundary" regime, lying outside the perturbative "soft-boundary" branch where the fixed/moving equivalence is strictly invertible via the Lambert W function.

Significance and Claims
The paper claims to provide a conceptual clarification of the role of boundaries in $T\bar{T}$-deformed theories. Its primary significance lies in:

• Decoupling Intrinsic Physics from Geometric Artifacts: It establishes that the intrinsic $T\bar{T}$ deformation is a boundary-localized flow governed by the displacement operator, regardless of whether the holographic realization manifests as a moving boundary (Type A) or a pinned boundary (Type B).
• Resolving Ambiguity: It distinguishes between the "fixed-moving boundary equivalence" (a field-theoretic statement about variational parameterizations) and "Type A vs. Type B" (inequivalent geometric realizations).
• Universality: By demonstrating that entanglement entropy is insensitive to the choice of holographic cutoff geometry, the work validates the intrinsic definition of the deformation and provides a robust framework for analyzing irrelevant deformations in boundary quantum field theories without introducing spurious boundary degrees of freedom.

The authors conclude that the deformation does not introduce new boundary dynamics but rather reorganizes existing BCFT data into a nonlinear functional of the displacement operator, fully determined by Ward identities.