Selecting Complex Extremal Surfaces with the Kontsevich--Segal--Witten Criterion

Technical Summary: Selecting Complex Extremal Surfaces with the Kontsevich–Segal–Witten Criterion

Problem Statement
In the context of AdS/CFT and dS/CFT, holographic observables associated with timelike subregions (such as timelike entanglement entropy) and observables in de Sitter space naturally involve complex extremal surfaces. While the Ryu–Takayanagi (RT) and Hubeny–Rangamani–Takayanagi (HRT) prescriptions relate spatial entanglement to real extremal surfaces in a semiclassical bulk, timelike correlations generally require evaluating extremal surfaces on nontrivial cycles of a complexified geometry. A central difficulty in this framework is the ambiguity of the integration contours for these complex saddles within the gravitational path integral. Direct determination via Picard–Lefschetz analysis is often intractable. Furthermore, it is unclear how to construct the corresponding complex bulk metrics and which specific complex geometries are physically admissible.

Methodology
The authors propose a constructive method to generate candidate complex bulk metrics from families of complex extremal surfaces. The core procedure involves:

1. Family Construction: Considering a continuous family of extremal surfaces in the bulk, parameterized by boundary subsystem parameters ($a_i$) and an intrinsic parameter ($\lambda$).
2. Complexification: Replacing the real intrinsic parameter $\lambda$ with a contour $\lambda(u)$ in the complex plane, where $u \in \mathbb{R}$. This defines a real-dimensional cycle in the complexified bulk.
3. Metric Pullback: Computing the pullback of the complexified bulk metric onto this real-dimensional cycle to obtain a complex family metric $g_{\alpha\beta}(q)$.
4. KSW Selection: Applying the Kontsevich–Segal–Witten (KSW) criterion as a consistency condition to constrain the contour $\lambda(u)$. The KSW criterion requires that for a complex metric diagonalized in a real basis, the sum of the absolute phases of its eigenvalues satisfies $\Theta[g] = \sum |\text{Arg } \Lambda_\gamma| \leq \pi$. The authors include marginal cases ($\Theta = \pi$) to encompass ordinary Lorentzian metrics.

Key Contributions and Results

• Universal Complex Metric in AdS$_3$:
  The authors demonstrate that Poincaré AdS$_3$, global AdS$_3$, and non-rotating BTZ black holes all induce the same local complex family metric:
  $$\frac{ds^2_C}{L_{\text{AdS}}^2} = d\lambda^2 + \cosh^2\lambda \, d\xi^2 - \sinh^2\lambda \, d\eta^2$$
  Applying the KSW criterion uniquely selects a specific three-piece contour in the complex $\lambda$ plane:

  • Spacelike Entanglement: The contour is purely real ($\lambda \in \mathbb{R}$), reproducing the standard Lorentzian AdS$_3$ geometry.
  • Timelike Entanglement: The contour consists of two horizontal branches ($\lambda = g \pm i\pi/2$) connected by a vertical segment on the imaginary axis ($\lambda = if$).
  • Geometric Interpretation: The resulting geometry is a piecewise-Lorentzian complex saddle. The horizontal branches correspond to the original Lorentzian sections (e.g., Poincaré or BTZ exteriors), while the middle vertical segment corresponds to a different real Lorentzian section (locally isometric to the interior of an AdS-Rindler wedge). The imaginary contribution to timelike entanglement entropy arises geometrically from the length of this middle timelike segment.

• Counterpart in dS$_3$:
  A parallel construction for global dS$_3$ yields a similar complex family metric but with reversed phase inequalities due to the positive cosmological constant. The KSW criterion uniquely selects a contour that decomposes the complexified dS$_3$ geometry into the original cosmological section and a middle Lorentzian patch (locally the static patch), connected across cosmological horizons.

• Higher-Dimensional Hyperbolic Surfaces (AdS$_4$):
  For AdS$_4$, the authors analyze extremal surfaces anchored to hyperbolic boundary regions. They show that the timelike hyperbolic family admits a complex metric analogous to the AdS$_3$ case. The KSW criterion selects the same three-piece contour structure, and the extra hyperbolic directions carry zero phase, preserving admissibility. This confirms that the timelike hyperbola and spacelike circle are different real sections of the same complexified quadratic family.

• KSW Obstruction for Planar Strips:
  The paper investigates planar strips in general dimensions $d$.

  • Spacelike Strips: Always reproduce a real Lorentzian AdS geometry, satisfying the KSW bound.
  • Timelike Strips: The behavior depends on the parity of the boundary dimension $d$.
    • In even dimensions (e.g., AdS$_5$), a straight real branch exists that saturates the KSW bound, though it extends to the Poincaré horizon rather than connecting the two boundary components.
    • In odd dimensions (e.g., AdS$_4$), the authors find a KSW obstruction. Perturbative analysis near the boundary shows that no contour within the general family can satisfy the KSW condition. Specifically, the phase sum $\Theta[g]$ exceeds $\pi$ at order $O(z)$ near the conformal boundary, regardless of the choice of subleading coefficients. This indicates that the induced complex geometry for timelike strips in odd dimensions violates the KSW bound.

Significance and Claims
The paper claims to provide a concrete realization of how complex bulk geometry may emerge from timelike entanglement, extending the paradigm that spacetime geometry is encoded in quantum entanglement. The primary significance lies in:

1. Selection Principle: The KSW criterion acts as a strong selection principle that uniquely determines the admissible integration contour for complex extremal surfaces in several non-trivial examples (AdS$_3$, dS$_3$, AdS$_4$ hyperbolic).
2. Geometric Origin of Imaginary Entropy: The construction provides a geometric explanation for the imaginary part of timelike entanglement entropy, attributing it to a segment of the extremal surface passing through a different real Lorentzian section of the complexified geometry.
3. Limitations and Obstructions: The work highlights the scope and limitations of the construction. The identification of configurations (like timelike strips in AdS$_4$) that violate the KSW bound suggests that not all timelike subsystems admit a KSW-allowable saddle within the proposed ansatz. This configuration dependence may encode part of the problem of selecting the correct complex saddle in the Lorentzian gravitational path integral.

The authors modestly state that their construction provides a possible selection principle and a set of candidate saddles, but it is not yet a first-principles derivation of the relevant integration cycle from the full gravitational path integral. They suggest that further investigation is needed to determine if these geometries can evaluate correlation functions at timelike separation and how they relate to existing real-time holography prescriptions.