CoChat

The DOI matches its Crossref record; 47 of the 47 references it cites were found in a registry or an index.

Holographic codes from hyperinvariant tensor networks

Open access
Steinberg, Matthew, Feld, Sebastian, Jahn, Alexander · 2023 · Nature Communications

Abstract

Holographic quantum-error correcting codes are models of bulk/boundary dualities such as the anti-de Sitter/conformal field theory (AdS/CFT) correspondence, where a higher-dimensional bulk geometry is associated with the code’s logical degrees of freedom. Previous discrete holographic codes based on tensor networks have reproduced the general code properties expected from continuum AdS/CFT, such as complementary recovery. However, the boundary states of such tensor networks typically do not exhibit the expected correlation functions of CFT boundary states. In this work, we show that a new class of exact holographic codes, extending the previously proposed hyperinvariant tensor networks into quantum codes, produce the correct boundary correlation functions. This approach yields a dictionary between logical states in the bulk and the critical renormalization group flow of boundary states. Furthermore, these codes exhibit a state-dependent breakdown of complementary recovery as expected from AdS/CFT under small quantum gravity corrections.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-023-42743-z

Cites

47 of 47 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Complementary recovery in holography. a In continuum AdS/CFT on slices at constant time t, a bipartition of the boundary CFT into two regions A and Ac is equivalent to a bipartition of the bulk into two entanglement wedges a and ac, separated by a Ryu–Takayanagi surface γA. Any bulk (field) operator ϕ(x) can be fully reconstructed on either A or Ac, but not both. b In the holographic tensor network code introduced here, a boundary bipartition leads to bulk wedges that are separated by a large residual region (white); an operator ϕ in this region cannot generally be reconstructed on either A and Ac.
Fig. 1 Complementary recovery in holography. a In continuum AdS/CFT on slices at constant time t, a bipartition of the boundary CFT into two regions A and Ac is equivalent to a bipartition of the bulk into two entanglement wedges a and ac, separated by a Ryu–Takayanagi surface γA. Any bulk (field) operator ϕ(x) can be fully reconstructed on either A or Ac, but not both. b In the holographic tensor network code introduced here, a boundary bipartition leads to bulk wedges that are separated by a large residual region (white); an operator ϕ in this region cannot generally be reconstructed on either A and Ac.CC BY · PubMed Central
Fig. 2 Bulk reconstruction in HaPPY codes. a A tensor network on a regular hyperbolic {4, 5} tiling, defining a map V from logical bulk legs (red) to physical boundary legs (black). b Conditions of the form (4) for perfect pentagon tensors, the choice of which makes V isometric and produces a HaPPY code. Dark-shaded tensors are conjugated. c The perfect tensor conditions define steps in a greedy algorithm for iteratively reconstructing the bulk from a boundary region A (bulk legs not drawn). d The greedy algorithm terminates at a minimal cut γA, a discrete Ryu–Takayanagi surface. Within the wedge a bounded by A ∪ γA, logical operators can be isometrically mapped to A.
Fig. 2 Bulk reconstruction in HaPPY codes. a A tensor network on a regular hyperbolic {4, 5} tiling, defining a map V from logical bulk legs (red) to physical boundary legs (black). b Conditions of the form (4) for perfect pentagon tensors, the choice of which makes V isometric and produces a HaPPY code. Dark-shaded tensors are conjugated. c The perfect tensor conditions define steps in a greedy algorithm for iteratively reconstructing the bulk from a boundary region A (bulk legs not drawn). d The greedy algorithm terminates at a minimal cut γA, a discrete Ryu–Takayanagi surface. Within the wedge a bounded by A ∪ γA, logical operators can be isometrically mapped to A.CC BY · PubMed Central
Fig. 3 Coarse-graining renormalization group step of a one-site operator. a In the {4, 5} HaPPY code, any single-site operator Oτ on layer τ of the tiling is a correctable error, and mapped to the identity on layer τ + 1. b In the {5, 4} HTN code, Oτ is generally mapped to a local S(Oτ)=Oτ+1≠1, hence allowing for non-trivial single-site lattice primary operator with Oτ+1∝Oτ. Tensors shaded in dark are conjugated, and red dots represent logical degrees of freedom, with legs suppressed.
Fig. 3 Coarse-graining renormalization group step of a one-site operator. a In the {4, 5} HaPPY code, any single-site operator Oτ on layer τ of the tiling is a correctable error, and mapped to the identity on layer τ + 1. b In the {5, 4} HTN code, Oτ is generally mapped to a local S(Oτ)=Oτ+1≠1, hence allowing for non-trivial single-site lattice primary operator with Oτ+1∝Oτ. Tensors shaded in dark are conjugated, and red dots represent logical degrees of freedom, with legs suppressed.CC BY · PubMed Central
Fig. 4 Construction of a hyperinvariant tensor network (HTN) code. a The original HTN construction from Ref. 22,23 for a {5, 4} hyperbolic tiling; vertices and edges are associated with A and B tensors, respectively. b The isometry constraints for one- and two-vertex combinations of A and B. With these constraints, each radial layer of tensors acts isometrically, in the manner of an RG transformation. c Implanting an additional leg on each A tensor yields a bulk Hilbert space, with the tensor network acting as a bulk-boundary map. d The updated isometry constraints (with logical legs in red) turn the RG step into a bulk reconstruction step. Despite the A tensors not being necessarily perfect as in the HaPPY code (Fig. 2), the tensor network acts as an exact bulk-to-boundary isometry.
Fig. 4 Construction of a hyperinvariant tensor network (HTN) code. a The original HTN construction from Ref. 22,23 for a {5, 4} hyperbolic tiling; vertices and edges are associated with A and B tensors, respectively. b The isometry constraints for one- and two-vertex combinations of A and B. With these constraints, each radial layer of tensors acts isometrically, in the manner of an RG transformation. c Implanting an additional leg on each A tensor yields a bulk Hilbert space, with the tensor network acting as a bulk-boundary map. d The updated isometry constraints (with logical legs in red) turn the RG step into a bulk reconstruction step. Despite the A tensors not being necessarily perfect as in the HaPPY code (Fig. 2), the tensor network acts as an exact bulk-to-boundary isometry.CC BY · PubMed Central
Fig. 5 Bulk reconstruction in HTN codes. a For perfect tensors, any index bipartition into regions A and Ac leads to a reconstruction of the logical index (red dot) on either A or Ac. b For the non-perfect tensors used in the HTN code, bipartitions exist for which neither A nor Ac are sufficient for reconstruction. c For a patch of the {5, 4} ququart HTN code with A^(′) and B tensors given by (9) and (12), we can show state-dependent reconstruction explicitly: For the given boundary regions A and Ac, the central bulk ququart cannot be state-independently reconstructed on either. But given local projections of the neighboring bulk ququart on eigenstates of the logical ququart Paulis X¯ and Z¯, it can be fully reconstructed on either A or Ac. While the RT surface γA bounding the reconstructible region changes, its length ∣γA∣ remains constant.
Fig. 5 Bulk reconstruction in HTN codes. a For perfect tensors, any index bipartition into regions A and Ac leads to a reconstruction of the logical index (red dot) on either A or Ac. b For the non-perfect tensors used in the HTN code, bipartitions exist for which neither A nor Ac are sufficient for reconstruction. c For a patch of the {5, 4} ququart HTN code with A^(′) and B tensors given by (9) and (12), we can show state-dependent reconstruction explicitly: For the given boundary regions A and Ac, the central bulk ququart cannot be state-independently reconstructed on either. But given local projections of the neighboring bulk ququart on eigenstates of the logical ququart Paulis X¯ and Z¯, it can be fully reconstructed on either A or Ac. While the RT surface γA bounding the reconstructible region changes, its length ∣γA∣ remains constant.CC BY · PubMed Central
Fig. 6 Effect of site erasures on bulk reconstruction. a In an HTN code, erasure of two boundary sites (black crosses) bisects the remaining boundary into two disjoint parts A1 and A2, with the reconstruction wedge of A1 ∪ A2 (shaded blue) being disjoint as well, separated by a strip-like region. b For three single-site erasures, the reconstruction wedge of A1 ∪ A2 ∪ A3 leaves out a larger region in the center of the bulk. Thus even small boundary operations can have an effect on information deep in the bulk for part of the codespace.
Fig. 6 Effect of site erasures on bulk reconstruction. a In an HTN code, erasure of two boundary sites (black crosses) bisects the remaining boundary into two disjoint parts A1 and A2, with the reconstruction wedge of A1 ∪ A2 (shaded blue) being disjoint as well, separated by a strip-like region. b For three single-site erasures, the reconstruction wedge of A1 ∪ A2 ∪ A3 leaves out a larger region in the center of the bulk. Thus even small boundary operations can have an effect on information deep in the bulk for part of the codespace.CC BY · PubMed Central
Fig. 7 Graphical proof of the state-dependent reconstruction of the central ququart in Fig. 5c. a The ququart Pauli operator X is represented by the logical operator X¯=XX²X³I. After pushing the Pauli operators acting on internal indices through the B tensors (green dots), which exchanges X ↔ Z, we apply the stabilizer (ZZ²ZI)³=Z³Z²Z³I on the left side and (Z¯)³=(Z³ZII)³=ZZ³II on the right, resulting in a physical operator represented only on boundary region A. b Similarly, we can push Z¯=Z³ZII to the left, apply (X¯)³=(XX²X³I)³=X³X²XI, and again arrive at a physical operator on A.
Fig. 7 Graphical proof of the state-dependent reconstruction of the central ququart in Fig. 5c. a The ququart Pauli operator X is represented by the logical operator X¯=XX²X³I. After pushing the Pauli operators acting on internal indices through the B tensors (green dots), which exchanges X ↔ Z, we apply the stabilizer (ZZ²ZI)³=Z³Z²Z³I on the left side and (Z¯)³=(Z³ZII)³=ZZ³II on the right, resulting in a physical operator represented only on boundary region A. b Similarly, we can push Z¯=Z³ZII to the left, apply (X¯)³=(XX²X³I)³=X³X²XI, and again arrive at a physical operator on A.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 47 references

Cited by

See all 35

Metadata

Authors
Steinberg, M., Feld, S., & Jahn, A.
Affiliations
  • QuTech (NL)
  • California Institute of Technology (US)
Corresponding author
Jahn, Alexander
Venue
Nature Communications
Published
2023-11-11
Volume
14
Issue
1
Pages
7314-7314
ISSN
2041-1723 (ISSN-L)
OpenAlex
W4388593305
Cited by
20
References
43
Citations by year
2026: 5 · 2025: 10 · 2024: 5
FWCI
1.46
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion, other-oa)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Research Repository (Delft University of Technology) (submittedVersion, cc-by)
Funders
Einstein Stiftung Berlin (ERUQD)
Topics
  • Physical Sciences › Physics and Astronomy › Nuclear and High Energy Physics › Black Holes and Theoretical Physics
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Quantum many-body systems
  • Physical Sciences › Computer Science › Artificial Intelligence › Quantum Computing Algorithms and Architecture
Keywords
AdS/CFT correspondence, boundary correlation functions, quantum gravity corrections, tensor networks

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Holographic codes from hyperinvariant tensor networks. https://app.cochat.ai/papers/10.1038/s41467-023-42743-z
The paper
Steinberg, M., Feld, S., & Jahn, A. (2023). Holographic codes from hyperinvariant tensor networks. Nature Communications. https://doi.org/10.1038/s41467-023-42743-z

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.