Ground State Entanglement in 1-Dimensional Translationally-Invariant Quantum Systems Sandy Irani Computer Science Department University of California,

Slides:



Advertisements
Similar presentations
One-Dimensional Spin Chains Are Hard Daniel Gottesman D. Aharonov, D. Gottesman, J. Kempe, arXiv: S. Irani, arXiv:
Advertisements

Lower Bounds for Additive Spanners, Emulators, and More David P. Woodruff MIT and Tsinghua University To appear in FOCS, 2006.
Local Hamiltonians in Quantum Computation Funding: Slovak Research and Development Agency, contract No. APVV , European Project QAP 2004-IST- FETPI-15848,
Part 2: Indirect interactions We assume that two quantum objects A and C interact only via another object (or objects) B. No other assumptions are made.
Equilibration and Unitary k- Designs Fernando G.S.L. Brandão UCL Joint work with Aram Harrow and Michal Horodecki arXiv: IMS, September 2013.
WSPD Applications.
13 May 2009Instructor: Tasneem Darwish1 University of Palestine Faculty of Applied Engineering and Urban Planning Software Engineering Department Introduction.
Preparing Topological States on a Quantum Computer Martin Schwarz (1), Kristan Temme (1), Frank Verstraete (1) Toby Cubitt (2), David Perez-Garcia (2)
Outline. Theorem For the two processor network, Bit C(Leader) = Bit C(MaxF) = 2[log 2 ((M + 2)/3.5)] and Bit C t (Leader) = Bit C t (MaxF) = 2[log 2 ((M.
MRA basic concepts Jyun-Ming Chen Spring Introduction MRA (multi- resolution analysis) –Construct a hierarchy of approximations to functions in.
Reflection Symmetry and Energy-Level Ordering of Frustrated Ladder Models Tigran Hakobyan Yerevan State University & Yerevan Physics Institute The extension.
Boris Altshuler Columbia University Anderson Localization against Adiabatic Quantum Computation Hari Krovi, Jérémie Roland NEC Laboratories America.
Entropy Rates of a Stochastic Process
No friction. No air resistance. Perfect Spring Two normal modes. Coupled Pendulums Weak spring Time Dependent Two State Problem Copyright – Michael D.
Chapter 9 Graph algorithms. Sample Graph Problems Path problems. Connectedness problems. Spanning tree problems.
Julia Kempe CNRS & LRI, Univ. of Paris-Sud Alexei Kitaev Caltech Oded Regev Tel-Aviv University FSTTCS, Chennai, December 18 th, 2004 The Complexity of.
1 Computing Nash Equilibrium Presenter: Yishay Mansour.
Validating Streaming XML Documents Luc Segoufin & Victor Vianu Presented by Harel Paz.
Chapter 9 Graph algorithms Lec 21 Dec 1, Sample Graph Problems Path problems. Connectedness problems. Spanning tree problems.
Theta Function Lecture 24: Apr 18. Error Detection Code Given a noisy channel, and a finite alphabet V, and certain pairs that can be confounded, the.
Introduction to Quantum Information Processing Lecture 4 Michele Mosca.
Expanders Eliyahu Kiperwasser. What is it? Expanders are graphs with no small cuts. The later gives several unique traits to such graph, such as: – High.
Quantum Mechanics from Classical Statistics. what is an atom ? quantum mechanics : isolated object quantum mechanics : isolated object quantum field theory.
Route Finding: A Quantum [Non] Algorithm Jason Clemons EECS 598 November 7, 2001.
MOHAMMAD IMRAN DEPARTMENT OF APPLIED SCIENCES JAHANGIRABAD EDUCATIONAL GROUP OF INSTITUTES.
Efficient Quantum State Tomography using the MERA in 1D critical system Presenter : Jong Yeon Lee (Undergraduate, Caltech)
Subdivision Analysis via JSR We already know the z-transform formulation of schemes: To check if the scheme generates a continuous limit curve ( the scheme.
Linear Algebra Chapter 4 Vector Spaces.
Molecular Information Content
Introduction to Tensor Network States Sukhwinder Singh Macquarie University (Sydney)
Theory of Computing Lecture 15 MAS 714 Hartmut Klauck.
Week 11 - Wednesday.  What did we talk about last time?  Graphs  Euler paths and tours.
1 CS104 : Discrete Structures Chapter V Graph Theory.
Mathematics of Networks (Cont)
By J. Burns and J. Pachl Based on a presentation by Irina Shapira and Julia Mosin Uniform Self-Stabilization 1 P0P0 P1P1 P2P2 P3P3 P4P4 P5P5.
Barriers in Hamiltonian Complexity Umesh V. Vazirani U.C. Berkeley.
Week 11 - Monday.  What did we talk about last time?  Binomial theorem and Pascal's triangle  Conditional probability  Bayes’ theorem.
H ij Entangle- ment flow multipartite systems [1] Numerically computed times assuming saturated rate equations, along with the lower bound (solid line)
1 Dorit Aharonov Hebrew Univ. & UC Berkeley Adiabatic Quantum Computation.
1 12/2/2015 MATH 224 – Discrete Mathematics Formally a graph is just a collection of unordered or ordered pairs, where for example, if {a,b} G if a, b.
The complexity of poly-gapped Hamiltonians (Extending Valiant-Vazirani Theorem to the probabilistic and quantum settings) Fernando G.S.L. Brandão joint.
Introduction to topological superconductivity and Majorana fermions
A Monomial matrix formalism to describe quantum many-body states Maarten Van den Nest Max Planck Institute for Quantum Optics Montreal, October 19 th 2011.
On Coding for Real-Time Streaming under Packet Erasures Derek Leong *#, Asma Qureshi *, and Tracey Ho * * California Institute of Technology, Pasadena,
Chapter 11 Resource Allocation by Mikhail Nesterenko “Distributed Algorithms” by Nancy A. Lynch.
GRAPHS. Graph Graph terminology: vertex, edge, adjacent, incident, degree, cycle, path, connected component, spanning tree Types of graphs: undirected,
Mathematical Tools of Quantum Mechanics
Physical Chemistry III (728342) The Schrödinger Equation
A counterexample for the graph area law conjecture Dorit Aharonov Aram Harrow Zeph Landau Daniel Nagaj Mario Szegedy Umesh Vazirani arXiv:
Approximation Algorithms Greedy Strategies. I hear, I forget. I learn, I remember. I do, I understand! 2 Max and Min  min f is equivalent to max –f.
Week 11 - Wednesday.  What did we talk about last time?  Graphs  Paths and circuits.
1 Objective To provide background material in support of topics in Digital Image Processing that are based on matrices and/or vectors. Review Matrices.
Mathematical Formulation of the Superposition Principle
Hamiltonian quantum computer in one dimension
Markov Chains Mixing Times Lecture 5
Hamiltonian cycle part
Lecture 18: Uniformity Testing Monotonicity Testing
Generalized DMRG with Tree Tensor Network
Enumerating Distances Using Spanners of Bounded Degree
Graph Theory.
Lecture on Linear Algebra
Computational Complexity of Quantum Systems
Graph Operations And Representation
Introduction Wireless Ad-Hoc Network
Shrödinger Equation.
Concept test 14.1 Is the function graph d below a possible wavefunction for an electron in a 1-D infinite square well between
Chapter 9 Graph algorithms
in collaboration with Andrew Doherty (UQ)
Computational approaches for quantum many-body systems
Introduction to topological superconductivity and Majorana fermions
Presentation transcript:

Ground State Entanglement in 1-Dimensional Translationally-Invariant Quantum Systems Sandy Irani Computer Science Department University of California, Irvine Visiting Institute for Quantum Information at Caltech arXiv:

Introduction What are the minimal set of properties a Hamiltonian must have in order for its ground state to have a high degree of entanglement? Do symmetries such as translational- invariance limit entanglement?

Introduction What are the minimal set of properties a Hamiltonian must have in order for its ground state to have a high degree of entanglement? Do symmetries such as translational- invariance limit entanglement? –One-dimensional systems. –Entropy of entanglement when traced down to a contiguous region in the 1D chain.

Previous Work Local Hamiltonian is QMA-complete, even for 1D systems. Adiabatic computation on 1D systems can perform universal quantum computation [AIGK07] Can these constructions be made translationally-invariant? –Translationally invariant modification that can be used for 1D universal adiabatic computation. [Nagaj-Wocjan, Janzin-Wocjan-Zhang] » Degenerate –1D Local Hamiltonian is QMA-complete, even what all two-particle terms are the same. [Kay] »Requires position-dependent 1-particle terms.

Translation-Invariance Can we make the 1D construction translationally-invariant with a Hamiltonian which has a non-degenerate ground state? If the system is described by a single Hamiltonian term applied to all pairs of particles (with bounded precision), how do we encode a circuit? –Show high entanglement in the ground state. –For 1D systems, previously known bounds, ground state entanglment entropy scales logarithmically with the region size.

Results Finite 1D chain: –Single Hamiltonian term H that operates on two particles of dimension 21, when applied to every neighboring pair in a finite chain of n particles: Unique ground state Spectral gap 1/poly(n) The entropy of entanglement of a region of size m on either end of the chain is W (min{m,n-m}).

Results Cycles and the Infinite Chain: –Family of translationally-invariant Hamiltonians {H n } for a cycle of nt 21-dimensional particles Spectral gap is 1/poly(n) For any state in the ground space of H n, and any m, there exist regions of size of m whose entanglement entropy is W (min{m,n}). –Entanglement bounds for a constant fraction of regions of size m. –Bounds hold in the limit as t tends towards infinity.

1D Area Law Upper bound on the entanglement entropy for any ground state of a 1D Hamiltonian H, independent of region size but exponentially dependent on 1/ D, where D is the spectral gap of H. [Hastings 07] Gottesman and Hastings: is the dependence on 1/ D tight? –Family of 1D Hamiltonians with unique ground state with regions whose entanglement entropy is W (poly(1/ D )). –Previously, best known such lower bound was W (log(1/ D )) arXiv: Construction present here gives a similar result.

Hamiltonian Construction Basics Type I terms (illegal pairs) –|ab><ab| –Energy penalty for: ….xxxabxxxxx…. Type II terms (transition rules) –½(|ab> <ab|) …xxxxabxxxx… …xxxxcdxxxx…

Hamiltonian Construction Basics Type I terms (illegal pairs) –|ab><ab| –Energy penalty for: ….xxxabxxxxx…. Type II terms (transition rules) –½(|ab> <ab|) …xxxxabxxxx… …xxxxcdxxxx… ab -> cd

Hamiltonian Construction Basics Ground State = Each contains no illegal pairs. [Kitaev02]

One Dimensional Hamiltonians Two types of states: –Control states:  –Passive States: W w E e U u Transition rules apply to control state and a state to the right or left. –May move control state to the left or right. –For example: --> [AGIK07] w  W

One Dimensional Hamiltonians Two types of states: –Control states:  –Passive States: W w E e U u Transition rules apply to control state and a state to the right or left. –May move control state to the left or right. –For example: --> [AGIK07] w  W w  eeWE … …

One Dimensional Hamiltonians Two types of states: –Control states:  –Passive States: W w E e U u Transition rules apply to control state and a state to the right or left. –May move control state to the left or right. –For example: --> [AGIK07] w  W w  eeWE … …

One Dimensional Hamiltonians Two types of states: –Control states:  –Passive States: W w E e U u Transition rules apply to control state and a state to the right or left. –May move control state to the left or right. –For example: --> [AGIK07] w  W w   W ee eeWE WE … …… …

Construction Overview W  >< e+e+ w U0U0 uE1E1 Circles: Single states Diamonds: Two-dimensional subsystems   Control states: Standard basis: specify state type for each site And then 0 or 1 for each 2D subsystem  

Construction Overview W ++ > < e+e+ E+E+ w u+u+ e+e+ u+u+ u+u+ e+e+ W E+E+ EPR Pairs W e+e+ E1E1 U0U0 u Ww Entangled states Unentangled states Waiting states

Target Ground State W >< e+e+ E+E+ U+U+ U+U+ W 

W >< e+e+ E+E+ U+U+ U+U+ W 

W >< e+e+ E+E+ U+U+ U+U+ W   W >< e+e+ E+E+ U+U+ W e+e+

W >< e+e+ E+E+ U+U+ U+U+ W   W >< e+e+ E+E+ U+U+ W e+e+

W >< e+e+ E+E+ U+U+ U+U+ W   W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+

W >< e+e+ E+E+ w U+U+ U+U+ W   W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+

W >< e+e+ E+E+ w U+U+ U+U+ W   W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w

W >< e+e+ E+E+ w U+U+ U+U+ W   W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w

W >< e+e+ E+E+ w U+U+ U+U+ W    W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+ w

W >< e+e+ E+E+ w U+U+ U+U+ W    W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+  E+E+

W >< e+e+ E+E+ w U+U+ U+U+ W    W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+  E+E+ >< e+e+ E+E+ u+u+ e+e+  E+E+ W

W >< e+e+ E+E+ w U+U+ U+U+ W    W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+  E+E+ >< e+e+ E+E+ u+u+ e+e+  E+E+ W >< e+e+ E+E+ e+e+  E+E+ W U+U+

W >< e+e+ E+E+ w U+U+ U+U+ W    W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+  E+E+ >< e+e+ E+E+ u+u+ e+e+  E+E+ W >< e+e+ E+E+ e+e+  E+E+ W U+U+

W >< e+e+ E+E+ w U+U+ U+U+ W    W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+  E+E+ >< e+e+ E+E+ u+u+ e+e+  E+E+ W >< e+e+ E+E+ e+e+  E+E+ W U+U+ >< e+e+ E+E+ e+e+ E+E+ W U+U+ 

W >< e+e+ E+E+ w U+U+ U+U+ W    W >< e+e+ E+E+ U+U+ W  W >< e+e+ E+E+ W u+u+ e+e+ e+e+  >< e+e+ E+E+ W u+u+ e+e+ w  >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+ w w >< e+e+ E+E+ u+u+ e+e+  E+E+ >< e+e+ E+E+ u+u+ e+e+  E+E+ W >< e+e+ E+E+ e+e+  E+E+ W U+U+ >< e+e+ E+E+ e+e+ E+E+ W U+U+  … >< e+e+ E+E+ e+e+ E+E+  e+e+ E+E+

First Round W >< w U+U+ U+U+ W  W >< e+e+ U+U+ W W >< e+e+ W u+u+ >< e+e+ W u+u+ w >< e+e+ u+u+ w w >< e+e+ u+u+ w W U+U+  W >< U+U+ U+U+ WW e+e+ u+u+ u+u+ u+u+ u+u+ u+u+  W W W W w w >< e+e+ u+u+ w u+u+ w     Special control states for the first round force the left end and the right end to agree that it is the first round. Will be used to check that state is symmetric about center.

The Hamiltonian…so far H trans = sum of terms from transition rules as applied to all neighboring pairs of particles. H legal = sum of terms from illegal pairs H = H trans + H legal

Well-Formed States A state in the standard basis is well-formed if it is of the form: –( + e ) …or any substring of these forms Can be checked by local checks

Well-Formed States Can not enforce by local checks that a legal state has a control state. – E.g., eeee……eee must be a legal state. If the state is bracketed ( a on the right end), then a legal state must be exactly on of the following three forms: –( + e )

Properties of Well-Formed States H trans + H legal is closed over the subspace spanned by well-formed states. –(All additional terms will be diagonal in standard basis). For each well-formed state, at most one transition rule applies in the forward direction and at most one transition rule applies in the reverse direction.

State Graph Nodes – set of all state in standard basis. Edges - directed edge from state A to state B if B can be obtained by applying one transition rule to A. –Well-formed states are disconnected from the rest of the graph. –State graph restricted to well-formed states form disjoint paths.

Hamiltonians Restricted to Paths

The Hamiltonian restricted to a single path If H legal has a non-zero entry, the minimum eigenvalue of H restricted to the subspace spanned by states in the path is W (1/l 3 ) which is W (1/n 6 ). [Kitaev 02] If H legal is all zero, the state which is the uniform superposition of states in the path has zero energy.

Bad States Will check for these states by showing that they evolve (via forward or backward application of the transition rules) to illegal states. w  > < eEue w ee w  > < EEuE w Ee  > < Eu w eeuu w  > < E0E0 w ee1e1 uu u w

How To Check for Bad States: An Example w > < eEue w ee 

w > < eEue w ee w > < eu w eeu w  

w > < eEue w ee w > < eu w eeu w w > < eu w eeu w   

w > < eEue w ee w > < eu w eeu w w > < eu w eeu w    W > < eU W eeU W ...

How To Check for Bad States: An Example w > < eEue w ee w > < eu w eeu w w > < eu w eeu w    W > < eU W eeU W  W > < U W eeU W U...

Need to Show: A path that starts with corresponds to a zero energy state for H. A path that does not start with contains a state that has an illegal pair OR does not have states that are bracketed. (Start with ). W >< U+U+ W U+U+ … … mm W >< U+U+ W U+U+ … … mm

Initializing Qubits H init = | >< | Penalty for state U - Ensures that ground state corresponds to a path whose initial state has qubits set to U-U- U-U- W >< U+U+ W U+U+ … … m m

Enforcing Bracketed States H bracket = I - | > < | H = 3( H trans + H legal ) + H init + H bracket –3( H trans + H legal ) term ensures there are no bracket terms in the middle. –H bracket term gives an energy benefit for having brackets at the end <<>>

Entropy of Entanglement … … O(n 2 )

Entropy of Entanglement … … O(n 2 ) A

Entropy of Entanglement … … O(n 2 ) A

Finite Cycle of size tn Change H legal so that the pair is allowed. Well-formed states look like: A sequence from a to a is a segment. H is closed on the set of well-formed states for a fixed set of segments. >< ><  W >< eE W ue  W >< e W u W >< U

Finite Cycle Cont. H = p(n)(H legal + H trans +H init ) + H size For p(n) large enough, using the Projection Lemma of Kempe-Kitaev-Regev, we can assume that the ground state of H is composed of tensor projects of ground states for finite chains. Ground state for finite chain of length l is Ground state for H will have form:

H size H size = (1/n)I - | >< | + (n-1)/T n [| > < | ] T l is the number of standard basis states in the support of the ground state for a segment of length l. if and only if l=n Otherwise >>  

Ground States for the Finite Cycle n orthogonal ground states, each a translation one site over. For any region size a constant fraction of the regions of that size have high entanglement. Superposition of all n states is translationally invariant and for every region size, all regions of that size have high entanglement. ><><><><>< n n nnn ……… … …

Open Problems Improve gap – lower bound on entropy as a function of 1/ D. For a given region size m, can we achieve high entropy for all regions of size m? Unique ground state for finite cycle and infinite chain? Can we achieve high entanglement entropy for all region sizes simultaneously for a translationally-invariant Hamiltonian?