Why does the TsT transformation remain conjectured rather than proven holographically equivalent?

# Why TsT Remains Conjectural

The TsT/$T\bar{T}$ correspondence sits in a peculiar middle ground: there is strong supporting evidence, but no complete holographic proof. Here's why the gap persists.

The Core Challenge

The conjecture claims [AG-2024.07-993] that a TsT transformation (a specific geometric deformation in string theory) on the bulk side corresponds to a $T\bar{T}$ deformation (a particular irrelevant operator) on the boundary CFT. The obstacle is that we lack a fully rigorous reconstruction of the bulk-boundary map at finite deformation.

In ordinary AdS/CFT, we can identify operators on one side with operators on the other because we work near the conformal point where perturbation theory is reliable. But $T\bar{T}$ is a *marginal* deformation—it changes the theory non-trivially at all scales. The field theory becomes increasingly intricate away from the conformal point, and the bulk geometry becomes non-standard. Mapping between them precisely requires solving questions about quantum reconstruction that remain open [AG-2025.06-1035].

What *Does* Work

Recent work has strengthened confidence without closing the proof:

• Asymptotic symmetries match: Researchers found that when you apply a state-dependent, non-local field redefinition, the worldsheet stress tensor and equations of motion of the TsT-transformed string transform identically to those of the original theory. Crucially, these results are "compatible with that of the $T\bar{T}$ deformed CFT$_2$" [AG-2024.07-993]—a non-trivial consistency check.

• Observables behave correctly: Wilson loops, entanglement entropy, and other holographic observables have been computed in TsT-deformed geometries and compared across different backgrounds [AG-2026.02-760]. They exhibit the expected deformation structure, which is a good sign but not a proof of equivalence.

• Extensions work: The $T\bar{T}$ deformation has been extended to include gravitational anomalies and black hole interiors [AG-2025.07-1044, AG-2025.03-802], and in each case the holographic dual (a deformed bulk geometry) can be constructed and checked for consistency. The fact that these generalizations hang together suggests the underlying correspondence is real.

• Complexity matches: The quantum complexity of $T\bar{T}$-deformed theories can be computed holographically and produces sensible results [AG-2024.08-204], another encouraging sign.

Why Proof Remains Hard

The real stumbling block is that holographic duality is intrinsically non-perturbative and non-constructive. We don't have an algorithmic way to invert the AdS/CFT map at strong coupling. The TsT transformation is well-defined geometrically, and the $T\bar{T}$ deformation is well-defined as an operator, but proving they're dual requires showing a one-to-one correspondence between observables—and that requires understanding how to extract bulk operators from the boundary theory when the coupling is strong and the deformation is large.

In essence: we've checked enough corners of the correspondence to be confident, but we haven't built the full bridge.