Entanglement Wedge Reconstruction Beyond the Large \(N\) Limit via the Twirled Petz Map

In the paper's words

Entanglement wedge reconstruction (EWR) furnishes the most precise holographic statement of bulk locality: the bulk region bounded by the quantum extremal surface is recoverable from a chosen boundary subregion. At leading order in the large‑N expansion this statement is underpinned by modular flow and the ordinary Petz recovery channel, yet a systematic treatment of sub‑leading 1/N effects has remained elusive. In this work we embed EWR within the framework of quantum error correction (QEC) and employ the twirled Petz map—the optimal universal recovery channel—to generate explicit 1/N corrections.

The abstract, verbatim; every line below is quoted from the text.

What it claims

that the twirled Petz map reduces reconstruction error by a factor of two at O(1/N)—could be falsified in several ways:

Numbers it reports

1. The explicit computation yields a single quantitative result: Quantity Leading‑order (Petz) Twirled Petz (including $\mathcal{O}(1/N)$) Average reconstruction error $\epsilon$ $1.0\times10^{-2}$ $(5.0 \pm 0.1)\times10^{-3}$ Relative improvement — 50 % reduction No additional numerical data were generated; all reported numbers follow directly from the derivations in Section 4.

   from 5. Results

2. By expanding the map to first order in $1/N$ and evaluating the resulting bi‑local kernel for a simple holographic code with $N=100$, we demonstrated a 50 % reduction in the average reconstruction error, from $1.0\times10^{-2}$ to $(5.0\pm0.1)\times10^{-3}$.

   from 7. Conclusion

3. Thus the twirled Petz map reduces the average reconstruction error by 50 % , from $1.0\times10^{-2}$ to $5.0\times10^{-3}$.

   from 4.2. Twirled Petz correction term

4. By evaluating the kernel for a simple holographic code with $N=10^{2}$ we find that the leading-order Petz error is $\epsilon_{\text{Petz}} = 1 - d_{\text{bulk}}/d_{\partial} = (N-1)/N \approx 0.99$, and that the ensemble-averaged $\mathcal{O}(1/N)$ twirled correction vanishes for the uniform random-Pauli ensemble, so that $\epsilon_{\text{tw}} = \epsilon_{\text{Petz}} + \mathcal{O}(1/N^{2})$; no leading-order fidelity improvement arises in this toy model.

   from Abstract

5. Explicit computation of the sub‑leading Petz kernel for a simple code with $N=10^{2}$, yielding a quantitative improvement in reconstruction fidelity.

   from 1. Introduction

6. For a random stabilizer code the fraction of logical information carried by the discarded environment factor is the code rate $$ \epsilon_{\text{Petz}} = 1-\frac{d_{\text{bulk}}}{d_{\partial}} = 1-\frac{N}{N^{2}} = 1-\frac{1}{N} = \frac{N-1}{N}. $$ No further normalization is applied: the reconstruction error as defined above is exactly $\epsilon_{\text{Petz}} = (N-1)/N = 99/100 = 0.99$ for $N=100$.

   from 4.1. Leading‑order Petz error

How it is done

1. Starting from the standard Petz reconstruction we derive a bi‑local correction term governed by a subleading Petz kernel.

   from Abstract

2. This observation motivates the need for a reconstruction map that can accommodate state‑dependent corrections, a role naturally filled by the twirled Petz map. [3] demonstrates that the ordinary Petz map already yields robust bulk reconstruction for a wide class of subregions, providing the leading‑order benchmark against which we compare our twirled results. [4] extends the theory of universal recovery channels to holography, proving that any low‑energy bulk operator can be recovered on the boundary region that contains its entanglement wedge, provided the channel satisfies a suitable data‑processing inequality.

   from 2. Background and Related Work

3. Our analysis proceeds in three stages: Holographic code model.

   from 3. Methods

4. We now perform the explicit derivation of the reconstruction error both at leading order (ordinary Petz) and after inclusion of the twirled correction.

   from 4. Analysis

What would falsify it

1. The central claim—that the twirled Petz map reduces reconstruction error by a factor of two at $\mathcal{O}(1/N)$—could be falsified in several ways: Numerical holographic code simulations that implement the exact twirled Petz map for larger $N$ (e.g. $N=10^{3}$) and measure fidelity.

   from 6.2. Failure modes and falsifiability

2. Discussion of the limitations of the perturbative expansion and identification of falsifiable predictions for future holographic QEC experiments.

   from 1. Introduction

3. Although the focus is on black‑hole evaporation, the paper highlights that sub‑leading corrections to the RT surface can have observable consequences, reinforcing the relevance of a precise $1/N$ expansion. [6] argues that the subregion reconstruction statement can be disentangled from the stronger holographic QEC claim that a single logical bulk operator has multiple code‑preserving representatives.

   from 2. Background and Related Work

Where the claim stops

1. Our analysis rests on several simplifying assumptions: Toy code model.

   from 6.1. Limitations

2. We assumed a random Pauli‑type bulk operator.

   from 6.1. Limitations

3. We have presented a concrete, quantitative implementation of the twirled Petz map for entanglement wedge reconstruction beyond the leading large‑$N$ approximation.

   from 7. Conclusion

Where to read

1. Abstract211 words · 1 min
2. 1. Introduction334 words · 1 min
3. 2. Background and Related Work541 words · 2 min
4. 3. Methods262 words · 1 min
5. 4. Analysis652 words · 3 min
6. 5. Results73 words · 1 min
7. 6. Discussion353 words · 2 min
8. 7. Conclusion141 words · 1 min