Non-Isometric Quantum Error Correction in Gravity Arjun Kar Department of Physics and Astronomy University of British Columbia

2025-05-02 0 0 799.79KB 43 页 10玖币
侵权投诉
Non-Isometric Quantum Error Correction in Gravity
Arjun Kar
Department of Physics and Astronomy, University of British Columbia,
6224 Agricultural Road, Vancouver, BC V6T 1Z1, Canada
E-mail: arjunkar@phas.ubc.ca
Abstract: We construct and study an ensemble of non-isometric error correcting codes in a
toy model of an evaporating black hole in two-dimensional dilaton gravity. In the preferred
bases of Euclidean path integral states in the bulk and Hamiltonian eigenstates in the bound-
ary, the encoding map is proportional to a linear transformation with independent complex
Gaussian random entries of zero mean and unit variance. Using measure concentration, we
show that the typical such code is very likely to preserve pairwise inner products in a set S
of states that can be subexponentially large in the microcanonical Hilbert space dimension
of the black hole. The size of this set also serves as an upper limit on the bulk effective field
theory Hilbert space dimension. Similar techniques are used to demonstrate the existence
of state-specific reconstructions of S-preserving code space unitary operators. State-specific
reconstructions on subspaces exist when they are expected to by entanglement wedge recon-
struction. We comment on relations to complexity theory and the breakdown of bulk effective
field theory.
arXiv:2210.13476v1 [hep-th] 24 Oct 2022
Contents
1 Introduction 1
2 Holographic dictionary 4
2.1 Canonical ensemble 5
2.2 End-of-the-world branes 7
2.3 Microcanonical ensemble 10
3 Overlap preservation 13
3.1 Deviation bound 14
3.2 On a relation to complexity theory 18
4 Bulk reconstruction 20
4.1 State-specific reconstruction 21
4.2 Entanglement wedge reconstruction 23
5 Discussion 27
5.1 The equilibrium basis 28
5.2 Relative subexponential states 29
5.3 Limitations on complexity 30
5.4 Fundamental averaging 32
5.5 Gravitational operators 33
5.6 Breakdowns of bulk effective field theory 34
1 Introduction
In holographic formulations of quantum gravity [1,2], semiclassical spacetime emerges from
the quantum information theory of the microscopic degrees of freedom [3,4]. The first hint
of this fundamental role played by information theory appears in the study of gravitational
entropy formulas [513] which express von Neumann entropies in quantum gravity in terms of
semiclassical quantities like hypersurface areas and the entropy of quantum fields propagating
on fixed backgrounds.
The arguments [9,1417] which lead to these formulas often make use of the Euclidean
gravitational path integral, a somewhat mysterious object which is semiclassical at first sight
but has been known nearly since its invention [18,19] to contain some amount of information
– 1 –
about the microscopic degrees of freedom in quantum gravity [20].1Understanding precisely
how much microscopic information is accessible via these Euclidean techniques is a very active
area of inquiry [2830] in light of recent advances concerning the application of some of the
more powerful of the aforementioned entropy formulas to the black hole information paradox
[1113,15,16,31].2
On their own, such entropy formulas and Euclidean gravity calculations have taught us
much about the structure of quantum information in quantum gravity (see [4246] for just a
few examples). But perhaps their deepest implication lies in their relationship with quantum
error correcting codes [4751]. Quantum error correcting codes were originally created to
allow for robust manipulation and transmission of quantum information [52], and their very
existence is somewhat surprising due to the no-cloning principle of quantum mechanics.
As a consequence of the entropy formulas, a quantum error correcting structure was dis-
covered in holography [47] which consists of a linear map Vthat encodes a bulk semiclassical
“code” Hilbert space Hbinto the microscopic “physical” boundary Hilbert space HB. The
basic idea is that semiclassical bulk states and operators on Hbare encoded redundantly by
Vin the microscopic Hilbert space HB, and losing access to some portions of HBdoes not
necessarily obstruct our ability to reconstruct bulk physics in some portions of Hb. The por-
tions in question are determined by the hypersurfaces appearing in the gravitational entropy
formulas we recalled above [45,49], and this phenomenon is sometimes called “entanglement
wedge reconstruction”.
Since its discovery in this fashion, the error correction structure in holography has come
to essentially supersede the holographic gravitational entropy formulas, as these formulas and
their generalizations are now understood as direct consequences of this structure [50,51]. As
such, the study of gravitational entropy is superseded by the study of the three fundamental
objects in the error correction structure: the code space Hb, the physical space HB, and
the encoding map V. As the behavior of gravitational entropy underpins our picture of the
emergent bulk spacetime, understanding the limits of semiclassical bulk physics is intimately
related to how far this error correction structure may be extended [51,5355].
The choice of Hbhas involved some degree of arbitrariness, and has generally been taken
to be a finite subspace of the bulk effective field theory Hilbert space, just large enough to
contain the perturbative semiclassical dynamics of interest [56]. On the other hand, HBis
almost always taken to be the complete microscopic Hilbert space of the holographic dual
1Lorentzian approaches to gravitational entropy have historically relied upon a string theoretic or holo-
graphic formulation of quantum gravity where the underlying degrees of freedom are explicitly known [2126].
However, recent progress in algebraic quantum field theory has supplied a new perspective on this issue, one
that may be as general and independent of stringy details as the Euclidean path integral [27].
2The generality and apparent utility of these formulas has even led to efforts to reproduce [3234] and
analyze [3541] them in gravitational theories which are not known to be holographic.
– 2 –
boundary theory, perhaps restricted to some fixed charge sector or microcanonical energy
window. The interpretation of Vin holography is as the “bulk-to-boundary map”, and in
Euclidean gravity it roughly corresponds to a rule for transforming bulk path integrals into
boundary path integrals.
Using the Euclidean gravitational path integral [16,4749] or tensor network models of
holography [57,58], it is possible to explicitly define the encoding map Vand reconstruct
bulk operators acting on the code space Hbby manipulating the physical degrees of freedom
in HB.3In these explicit situations, we will loosely refer to Vas the holographic dictionary,
despite its apparent difference from the original notion of the holographic dictionary where
(noting the lack of a non-perturbatively defined bulk theory) Vwas instead thought of as
an isomorphism from a bulk string theory Hilbert space to a boundary gauge theory Hilbert
space [2,59,60].
The standard situation in quantum error correction is for the encoding map Vto be an
isometry. This means Vsatisfies VV=Ib, where Ibis the identity on Hb. In some toy
models and many holographic scenarios, the dictionary Vreally is an isometry and leads
to an approximate quantum error correcting code between the bulk and boundary where all
states in Hblook roughly the same from a semiclassical perspective.4
However, extensions of the standard error correction structure are necessary and indeed
sufficient to understand more precise notions of information in bulk subregions [53,54], choices
of Hbcontaining states with highly dissimilar bulk entanglement structure [17], and even
certain extreme situations motivated by the interiors of evaporating or old black holes where
Vcan be arbitrarily far from an isometry [51,55,70,71].5Of these extensions to standard
error correction, the non-isometric extension is the least understood and the most relevant
for situations with strong gravitational effects. In the most detailed study to date of this non-
isometric error correction [55], the model under consideration was a tensor network model
similar in spirit to [58], with no obvious connection to gravity.
As was already explained in [55], the tensor network model has several drawbacks as a
model of gravity. For example, Lorentz and diffeomorphism invariance are hard to under-
3Other bulk reconstruction techniques [2,5964] make use of the causal structure of spacetime and provide
only indirect access to V, often without an explicit choice of code space Hb. This subtlety led to several puzzles
concerning the structure of effective field theory within holography, and unraveling these issues led directly to
the quantum error correction ideas we have been reviewing [47,56]. Modular theory has also been employed
as a reconstruction technique [65] which may be versatile enough to handle choices of Hbwhich include bulk
subregions that are separated from HBby a horizon [6669].
4Technically, this means that bulk subregions are encoded in boundary subregions for every state in Hb
in the same manner. This is sometimes called “complementary recovery” in error correction or “subregion
duality” in holography.
5In the non-isometric situation, the map Vmay annihilate some states in Hb. These null states have been
related to a sort of large diffeomorphism invariance [7274].
– 3 –
stand in such models, and the quantity which plays the role of the microcanonical black hole
Hilbert space dimension has no obvious connection to a geometric area as expected by the
Bekenstein-Hawking formula. On the other hand, the Euclidean path integral naturally allows
for manifest diffeomorphism invariance, and the entropy of the black hole is clearly related to
the horizon area. Indeed, computations in Euclidean gravity were the original justification for
the Bekenstein-Hawking formula itself. In view of these facts, it is necessary to understand
the extent to which the results obtained by tensor network analysis in [55] may be carried
over to a real gravitational theory defined using the Euclidean path integral.
The purpose of this paper is to construct and study the dictionary Vas a highly non-
isometric quantum error correcting code in a toy model of an evaporating black hole in
two-dimensional Euclidean dilaton gravity. We combine the Euclidean gravity techniques
which allow for a resolution of the information paradox [16] with the notion of non-isometric
error correction introduced in [55]. We find gravitational analogues of many of the results
of [55], with interesting differences in details. In particular, our analogue (3.26) of the main
theorem of [55] dealing with changes in semiclassical state overlaps under the dictionary is
strengthened by differences in measure concentration between the Haar ensemble and the
complex Gaussian ensemble, which enters our gravitational code. This strengthening carries
over to the bulk reconstruction analyses in e.g. (4.5). Moreover, the derivations themselves
are in fact a bit simpler than in [55], which is surprising as gravitational theories are generally
more complicated than Haar random unitary analyses.
Four sections follow. In Section 2, we define the dictionary Vin two-dimensional dilaton
gravity and briefly review how the model relates to evaporating black holes. In Section 3,
we prove the gravitational analogue of the fundamental theorem of [55], which allows an esti-
mation of how many states in the code Hilbert space may have preserved overlaps under the
dictionary V. In Section 4, we describe the reconstruction of code operators in a necessarily
state-specific manner and verify consistency of our results with entanglement wedge recon-
struction. We conclude in Section 5with a discussion of relations to complexity, fundamental
averaging, extensions to gravitational bulk operators, and the breakdown of bulk effective
field theory.
2 Holographic dictionary
We will define the holographic dictionary Vusing the Euclidean gravitational path inte-
gral. Our discussion will apply quite generally to asymptotically anti-de Sitter (AdS) two-
dimensional dilaton gravity theories, but for concreteness we will also spell out the details for
a particular theory: Jackiw-Teitelboim (JT) gravity [75,76], which is a theory of the metric
– 4 –
摘要:

Non-IsometricQuantumErrorCorrectioninGravityArjunKarDepartmentofPhysicsandAstronomy,UniversityofBritishColumbia,6224AgriculturalRoad,Vancouver,BCV6T1Z1,CanadaE-mail:arjunkar@phas.ubc.caAbstract:Weconstructandstudyanensembleofnon-isometricerrorcorrectingcodesinatoymodelofanevaporatingblackholeintwo-d...

展开>> 收起<<
Non-Isometric Quantum Error Correction in Gravity Arjun Kar Department of Physics and Astronomy University of British Columbia.pdf

共43页,预览5页

还剩页未读, 继续阅读

声明:本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知玖贝云文库,我们立即给予删除!
分类:图书资源 价格:10玖币 属性:43 页 大小:799.79KB 格式:PDF 时间:2025-05-02

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务
/ 43
客服
关注