How does the dimension-dependence in de Sitter/CFT compare to the well-established AdS case?

# How dS/CFT Dimension-Dependence Differs from AdS/CFT

The AdS/CFT correspondence is dimension-independent in a fundamental sense: a *d*-dimensional CFT dual lives at the boundary of (*d*+1)-dimensional anti–de Sitter space, and the holographic dictionary (bulk operators ↔ boundary operators, bulk reconstruction, etc.) works the same way regardless of *d*. By contrast, de Sitter/CFT exhibits dimension-dependence in ways that reveal deep asymmetries between the two spacetimes.

The Core Structural Difference

In AdS/CFT, the boundary is *spacelike and timelike* — it's a rigid conformal background where the CFT lives. In dS/CFT, the putative dual resides at *future infinity* — a null surface — and the bulk wavefunction is dual to the boundary partition function rather than to correlation functions [AG-2024.07-041]. This topological difference means that standard CFT machinery (unitarity, OPE structure, bulk reconstruction) does not transfer straightforwardly.

Dimension-Specific Obstacles

Three concrete dimension-dependent problems emerge:

1. Boundary Interpretation: When you take a Wheeler-DeWitt quantum state (the bulk wavefunction) to the asymptotic boundary in de Sitter, the meaning of the dual CFT state is fundamentally unclear — there's no natural analogue of the Dirichlet boundary condition that anchors AdS/CFT [AG-2024.03-503].

2. Bulk Reconstruction and Subregion Duality: In AdS/CFT, a subregion of the boundary CFT encodes a subregion of the bulk interior (the *Ryu-Takayanagi* mechanism). This fails in de Sitter; there's no established way to recover bulk spatial localization from boundary data [AG-2024.03-503], and the CFT conjugation operation differs markedly — you cannot extract de Sitter Green's functions by simply Wick-rotating AdS results [AG-2024.05-451].

3. Non-Unitarity vs. Unitarity: In AdS/CFT, the boundary CFT is unitary (positive probabilities, energy bounded below). In dS/CFT, the boundary theory is *non-unitary*, and unitarity is rescued only at the level of the *spectral density* of correlators — a non-perturbative condition [AG-2026.03-098]. This shift from operator-level unitarity to correlator-level unitarity is dimension-independent in principle, but its physical implications depend on the operator spectrum, which scales with *d*.

Symmetry Complication at Low Dimensions

The distinction between *scale invariance* and *conformal invariance* becomes acute in low-dimensional dS/CFT. In AdS/CFT, conformal invariance (which includes boosts) is automatic. In de Sitter, scale invariance can emerge without full conformal symmetry — a scenario realized in modified gravity theories but not in Einstein gravity [AG-2026.02-334]. This gap widens as dimension decreases; in *d* = 2, the situation is especially delicate, and the CFT's Hamiltonian acquires an SL(2,ℝ) deformation that has no AdS counterpart [AG-2025.12-833].

Low-Dimensional Reductions Illuminate the Difference

Two-dimensional de Sitter (dS₂) is the cleanest laboratory. Recent work shows that dS₂/CFT reduces to a *deformed* CFT whose Hamiltonian is modified (not the standard Virasoro action) [AG-2025.12-833], and the thermodynamic signature is a sign flip in heat capacity at fixed deformation strength — a feature wholly absent in AdS [AG-2024.10-462]. These low-dimensional constructions reveal that dS/CFT is not merely AdS/CFT with opposite curvature; it is a fundamentally different holographic principle.

---

In short: AdS/CFT is "dimension-universal" — the correspondence and its dictionary work identically for all *d*. dS/CFT is "dimension-sensitive" — boundary interpretation, bulk reconstruction, and unitarity structure all shift as *d* varies, and they do so in ways orthogonal to the AdS playbook.