Exact Toy-Code Reconstruction Theorems

Exact holographic toy codes prove exact recovery statements for finite-dimensional encodings with declared erasures and access structures. They can realize redundant logical operators and minimal-surface-like entropy formulas. Their exactness does not include continuum Type-III algebras, gravitational dressing, backreaction, a dynamical area term, string corrections, or finite-$N$ AdS/CFT.

Required background. Error Models, Codes, and Recovery Conditions supplies the Knill–Laflamme criterion. Entanglement Structure and Tensor-Network Ansätze supplies tensor-network notation.

Helpful background. Complementary Recovery and Cleaning Relations gives the access-structure theorem. Tensor-Network Models of Holographic Entanglement Geometry explains the geometric analogy. Erasure, Subregions, and Correctability and Tensor Networks as Encoding Maps: Scope and Limits supply the QEC interpretation and its limits.

A three-qutrit perfect-tensor code

Let all arithmetic be modulo $3$ and encode one logical qutrit as

$$V|i\rangle =\frac{1}{\sqrt3}\sum_{k=0}^{2} |k,\,k+i,\,k+2i\rangle, \qquad i=0,1,2.$$

The three codewords are orthonormal. Consider erasure of physical qutrit 1. For matrix units $|i\rangle\langle j|$, tracing out qutrits 2 and 3 gives

$$\begin{aligned} \operatorname{Tr}_{23} \left(V|i\rangle\langle j|V^\dagger\right) &=\frac13\sum_{k,k'}|k\rangle\langle k'| \delta_{k+i,k'+j}\delta_{k+2i,k'+2j}\\ &=\delta_{ij}\,\frac{I_3}{3}. \end{aligned}$$

Subtracting the two congruences forces $i=j$ and then $k=k'$. Thus the erased qutrit contains no logical information. The same calculation holds for either other single-qutrit erasure by symmetry. Equivalently, for every pair of errors $E_a,E_b$ supported on one erased qutrit,

$$P E_a^\dagger E_b P=c_{ab}P,$$

so the Knill–Laflamme condition holds exactly. Any two physical qutrits can recover the logical qutrit; no one-qutrit set can. This is the $(2,3)$ quantum secret-sharing access structure Cleve, Gottesman, and Lo 1999.

An explicit decoder

Suppose qutrit 1 is erased and qutrits 2 and 3 are received with values $b=k+i$ and $c=k+2i$. Then

$$i=c-b, \qquad k=2b-c.$$

A reversible linear circuit maps $|b,c\rangle$ to $|k,i\rangle$; discarding the uniformly mixed $k$ register recovers the logical state, including its coherences. The decoder is exact for every reference system entangled with the logical qutrit, not only for basis states.

Perfect tensors tile this mechanism into the HaPPY network. Greedy tensor cancellation produces an exact reconstructible region and an RT-like cut entropy for the discrete code Pastawski et al. 2015. Random tensor networks reproduce related entropy behavior with high probability in a large bond-dimension limit, which is a different, probabilistic statement Hayden et al. 2016.

First application

Treat the logical qutrit as a toy bulk degree of freedom and the three physical qutrits as boundary shares. Erasing share 1 is a boundary-subregion noise model. The calculation above proves that the complementary channel to share 1 is constant and that shares 2 and 3 recover every logical operator. The logical clock and shift operators therefore have exact representatives on each two-share authorized region.

The result is stronger than a leading-$N$ calculation but narrower in physical scope: it is a theorem about the displayed $V$, finite Hilbert spaces, and one-share erasures.

Adversarial control

Replace a qutrit by a continuum local algebra, impose a gravitational Gauss law, let the tensor depend on the logical state through backreaction, or perturb $V$ so that $V^\dagger V\neq I$. The trace calculation no longer proves a constant complementary channel. Even a perturbation of norm $\delta$ turns exact correction into a norm-dependent approximate statement.

The same refusal applies to a random tensor: a high-probability entropy formula is not an exact operator reconstruction theorem for every realization. No step in the finite proof supplies $G_N$, $\alpha'$, a KK spectrum, a dressing prescription, or a nonperturbative CFT dictionary.

Regime, evidence ceiling, and handoff

There is no large-$N$ expansion inside the three-qutrit theorem. Its control parameters are finite dimension, exact isometry, and the specified erasure. A network adds bond dimension and graph geometry, but those are model parameters rather than a derivation of $N$, $g_s$, $\alpha'$, or curvature in a top-down compactification.

The evidence ceiling is exact finite-dimensional QEC plus a geometric analogy. Continue to Leading Semiclassical JLMS and Code-Subspace Claims to see what gravity actually supplies and to Non-Isometric Encoding Proposals for controlled departures from isometry.

The chapter overview contains the structure diagram and validity and failure diagram. They are embedded there once so that their shared chapter-level context is not repeated on every article.

For the chapter-wide comparison of assumptions, counterevidence, falsifiers, and claim ceilings, see the claim-domain table.

References

• Cleve, R., Gottesman, D., and Lo, H.-K. (1999), “How to Share a Quantum Secret,” Physical Review Letters 83, 648–651. DOI; arXiv:quant-ph/9901025.
• Hayden, P., Nezami, S., Qi, X.-L., Thomas, N., Walter, M., and Yang, Z. (2016), “Holographic Duality from Random Tensor Networks,” Journal of High Energy Physics 2016(11), 009. DOI; arXiv:1601.01694.
• Pastawski, F., Yoshida, B., Harlow, D., and Preskill, J. (2015), “Holographic Quantum Error-Correcting Codes: Toy Models for the Bulk/Boundary Correspondence,” Journal of High Energy Physics 2015(06), 149. DOI; arXiv:1503.06237.