The Classification of Clifford Gates over Qubits

Similar documents
Ph 219b/CS 219b. Exercises Due: Wednesday 20 November 2013

Quantum gate. Contents. Commonly used gates

QUANTUM COMPUTATION. Exercise sheet 1. Ashley Montanaro, University of Bristol H Z U = 1 2

Quantum Information & Quantum Computing

Fourier Sampling & Simon s Algorithm

92 CHAPTER III. QUANTUM COMPUTATION. Figure III.11: Diagram for swap (from NC).

Ph 219b/CS 219b. Exercises Due: Wednesday 21 November 2018 H = 1 ( ) 1 1. in quantum circuit notation, we denote the Hadamard gate as

C/CS/Phys 191 Quantum Gates and Universality 9/22/05 Fall 2005 Lecture 8. a b b d. w. Therefore, U preserves norms and angles (up to sign).

Quantum Information & Quantum Computing

Boolean State Transformation

Quantum Computation. Michael A. Nielsen. University of Queensland

Problem Set # 6 Solutions

1 Readings. 2 Unitary Operators. C/CS/Phys C191 Unitaries and Quantum Gates 9/22/09 Fall 2009 Lecture 8

Simulating classical circuits with quantum circuits. The complexity class Reversible-P. Universal gate set for reversible circuits P = Reversible-P

Lecture 6: Quantum error correction and quantum capacity

C/CS/Phys C191 Quantum Gates, Universality and Solovay-Kitaev 9/25/07 Fall 2007 Lecture 9

Introduction to Quantum Computing

Lecture 3: Hilbert spaces, tensor products

5 Symmetries and point group in a nut shell

UNDERSTANDING THE COST OF GROVER'S ALGORITHM FOR FINDING A SECRET KEY

Hilbert Space, Entanglement, Quantum Gates, Bell States, Superdense Coding.

b) (5 points) Give a simple quantum circuit that transforms the state

Applying Grover s algorithm to AES: quantum resource estimates

Unitary Dynamics and Quantum Circuits

. Here we are using the standard inner-product over C k to define orthogonality. Recall that the inner-product of two vectors φ = i α i.

Introduction to Quantum Information Processing QIC 710 / CS 768 / PH 767 / CO 681 / AM 871

Projects about Quantum adder circuits Final examination June 2018 Quirk Simulator

Quantum Computation CMU BB, Fall Week 6 work: Oct. 11 Oct hour week Obligatory problems are marked with [ ]

Single qubit + CNOT gates

Teleportation-based approaches to universal quantum computation with single-qubit measurements

6.896 Quantum Complexity Theory September 9, Lecture 2

Quantum Error Correction and Fault Tolerance. Classical Repetition Code. Quantum Errors. Barriers to Quantum Error Correction

Quantum computing! quantum gates! Fisica dell Energia!

CS286.2 Lecture 8: A variant of QPCP for multiplayer entangled games

More advanced codes 0 1 ( , 1 1 (

2.2 Classical circuit model of computation

Introduction to Group Theory

Advanced Cryptography Quantum Algorithms Christophe Petit

Chapter 10. Quantum algorithms

Some notes on Coxeter groups

Analyzing Quantum Circuits Via Polynomials

Bounds on Quantum codes

Quantum Computing. Quantum Computing. Sushain Cherivirala. Bits and Qubits

Quantum information and quantum computing

Lecture 2: From Classical to Quantum Model of Computation

A Study of Topological Quantum Error Correcting Codes Part I: From Classical to Quantum ECCs

Quantum Computing: From Circuit To Architecture

Complex numbers: a quick review. Chapter 10. Quantum algorithms. Definition: where i = 1. Polar form of z = a + b i is z = re iθ, where

Computational Methods in Finite Geometry

Simulation of quantum computers with probabilistic models

Fault-Tolerant Universality from Fault-Tolerant Stabilizer Operations and Noisy Ancillas

(IN)EQUIVALENCE OF COLOR CODE AND TORIC CODE. Aleksander Kubica, B. Yoshida, F. Pastawski

Grover s algorithm. We want to find aa. Search in an unordered database. QC oracle (as usual) Usual trick

Principles of Quantum Mechanics Pt. 2

6. Quantum error correcting codes

A Novel Synthesis Algorithm for Reversible Circuits

M. Grassl, Grundlagen der Quantenfehlerkorrektur, Universität Innsbruck, WS 2007/ , Folie 127

6.080/6.089 GITCS May 6-8, Lecture 22/23. α 0 + β 1. α 2 + β 2 = 1

On Projective Planes

Extended Superposed Quantum State Initialization Using Disjoint Prime Implicants

Chapter 11 Evolution by Permutation

Introduction to Quantum Algorithms Part I: Quantum Gates and Simon s Algorithm

Quantum Error Correcting Codes and Quantum Cryptography. Peter Shor M.I.T. Cambridge, MA 02139

SUPERDENSE CODING AND QUANTUM TELEPORTATION

QUANTUM COMPUTING 10.

Lecture 10: A (Brief) Introduction to Group Theory (See Chapter 3.13 in Boas, 3rd Edition)

Quantum Gates, Circuits & Teleportation

Fourier analysis of boolean functions in quantum computation

X row 1 X row 2, X row 2 X row 3, Z col 1 Z col 2, Z col 2 Z col 3,

Quantum Algorithms via Linear Algebra

(1.1) In particular, ψ( q 1, m 1 ; ; q N, m N ) 2 is the probability to find the first particle

Daniel J. Bernstein University of Illinois at Chicago. means an algorithm that a quantum computer can run.

Bounded depth quantum circuits

Quantum Computing: Foundations to Frontier Fall Lecture 3

Exercise Sheet 3 - Solutions

arxiv: v2 [quant-ph] 12 Aug 2008

sparse codes from quantum circuits

arxiv: v2 [quant-ph] 5 Oct 2017

Introduction to Quantum Computing

Quantum Computing Lecture 6. Quantum Search

CHAPTER 12 Boolean Algebra

THE ABC OF COLOR CODES

Actively secure two-party evaluation of any quantum operation

IBM quantum experience: Experimental implementations, scope, and limitations

2.0 Basic Elements of a Quantum Information Processor. 2.1 Classical information processing The carrier of information

Algebraic Properties and Panconnectivity of Folded Hypercubes

Richard Cleve David R. Cheriton School of Computer Science Institute for Quantum Computing University of Waterloo

5.5. Representations. Phys520.nb Definition N is called the dimensions of the representations The trivial presentation

Quantum Information & Quantum Computation

Quantum Computing. 6. Quantum Computer Architecture 7. Quantum Computers and Complexity

Seminar 1. Introduction to Quantum Computing

Assignment 3. A tutorial on the applications of discrete groups.

Section 4.1 Switching Algebra Symmetric Functions

On The Complexity of Quantum Circuit Manipulation

Analyzing Quantum Circuits Via Polynomials

Quantum parity algorithms as oracle calls, and application in Grover Database search

Theory of fault-tolerant quantum computation

arxiv:quant-ph/ v1 16 Jul 2003

CHAPTER 2 AN ALGORITHM FOR OPTIMIZATION OF QUANTUM COST. 2.1 Introduction

Ancilla-driven universal quantum computing

Transcription:

The Classification of Clifford Gates over Qubits Daniel Grier, Luke Schaeffer MIT

Introduction Section 1 Introduction

Classification Introduction Idea Study universality by finding non-universal gate sets.

Classification Introduction Idea Study universality by finding non-universal gate sets. Each non-universal set has some property (an invariant) which is an obstacle to universality. Sometimes non-universal sets are interesting in their own right. E.g., Clifford gates.

Classification Introduction Idea Study universality by finding non-universal gate sets. Each non-universal set has some property (an invariant) which is an obstacle to universality. Sometimes non-universal sets are interesting in their own right. E.g., Clifford gates. Definition Let G denote the set of unitaries constructible as circuits with gates from G.

Classification Introduction Idea Study universality by finding non-universal gate sets. Each non-universal set has some property (an invariant) which is an obstacle to universality. Sometimes non-universal sets are interesting in their own right. E.g., Clifford gates. Definition Let G denote the set of unitaries constructible as circuits with gates from G. Goal: Characterize all possible sets G.

Warm up: Reversible Gates Introduction Definition A gate is reversible if it maps each classical input to some classical output. E.g., CNOT, the Toffoli gate, CSWAP.

Warm up: Reversible Gates Introduction Definition A gate is reversible if it maps each classical input to some classical output. E.g., CNOT, the Toffoli gate, CSWAP. Completely classified (Aaronson, Grier, S. 2017) in a classical circuit model. We will consider a simplified version under a quantum circuit model.

Reversible Gate Classification Introduction Theorem Given a collection G of reversible gates, G is one of the following six sets: Toffoli, Fredkin, CNOT, T 4, (where T 4 is yet to be defined) NOT, or SWAP.

Introduction Toffoli CNOT T 4 Fredkin NOT

Invariants Introduction Observe that CNOT satisfies the following property. Affine Invariant Every output bit is an affine function (modulo 2) of the input bits.

Invariants Introduction Observe that CNOT satisfies the following property. Affine Invariant Every output bit is an affine function (modulo 2) of the input bits. Also, if all gates in a circuit satisfy this property, so does the entire circuit. (This is why it is an invariant)

Invariants Introduction Observe that CNOT satisfies the following property. Affine Invariant Every output bit is an affine function (modulo 2) of the input bits. Also, if all gates in a circuit satisfy this property, so does the entire circuit. (This is why it is an invariant) Similarly, Fredkin satisfies the following property. Conservativity Invariant The Hamming weight of the input bits is the same the Hamming weight of the output bits. Once again, if all gates have this property then so does the circuit.

Obstacles to Universality Introduction If all gates are affine, then the gate set is not universal. If all gates are conservative, then the gate set is not universal.

Obstacles to Universality Introduction If all gates are affine, then the gate set is not universal. If all gates are conservative, then the gate set is not universal. On the other hand, the classification shows us that these are the only obstacles. If at least one gate is not affine, and at least one gate is not conservative, then the gate set is universal.

Clifford Group Introduction Recall the Pauli matrices ( ) ( ) 1 0 0 1 I =, X =, 0 1 1 0 ( ) ( ) 0 i 1 0 Y =, Z =. i 0 0 1 The Pauli group on n qubits is P n = {±1, ±i} {I, X, Y, Z} n. The Clifford group on n qubits is the set of unitaries that normalize the Pauli group. C n = {U : UP n U = P n }

Introduction For example, CNOT C 2. CNOT (X I ) CNOT = X X CNOT (I X ) CNOT = I X CNOT (Z I ) CNOT = Z I CNOT (I Z) CNOT = Z Z

Introduction For example, CNOT C 2. Similarly, H = 1 2 ( 1 1 1 1 ), S = ( 1 0 0 i ) C 1. CNOT (X I ) CNOT = X X CNOT (I X ) CNOT = I X CNOT (Z I ) CNOT = Z I CNOT (I Z) CNOT = Z Z HXH = Z SXS = Y HZH = X SZS = Z

Introduction For example, CNOT C 2. Similarly, H = 1 2 ( 1 1 1 1 ), S = ( 1 0 0 i ) C 1. CNOT (X I ) CNOT = X X CNOT (I X ) CNOT = I X CNOT (Z I ) CNOT = Z I CNOT (I Z) CNOT = Z Z HXH = Z SXS = Y HZH = X SZS = Z Fact: CNOT, H, S = n C n.

Introduction Classification of Clifford Gates Theorem Given a collection G of Clifford gates, G is one of 57 possible classes.

Introduction Classification of Clifford Gates Theorem Given a collection G of Clifford gates, G is one of 57 possible classes. 57 sets is too many to list or draw on one slide. 30 classes are trivial, generated by single qubit gates. 27 more interesting classes

Introduction P + Γ θ +++ θ + θ + θ + P + R X P + R Y P + R Z P R X R Y R Z X + θ YZ Y + θ XZ Z + θ XY X Y Z Γ +++ Γ + Γ + Γ + θ Y +Z θ Y Z θ X +Z θ X Z θ X +Y θ X Y

Introduction ALL C(Y, X ) + P + R X T4 + P + Γ C(Z, Y ) + P + R Y C(X, Z) + P + R Z C(X, Z) + P T4 + P + R X C(Y, X ) + P T4 + P + R Y C(Z, Y ) + P T4 + P + R Z T4 + P C(X, X ) + P + R X C(Y, Y ) + P + R Y C(Z, Z) + P + R Z C(X, X ) + P C(X, X ) + R X C(X, X ) + X + θ YZ C(Y, Y ) + P C(Y, Y ) + R Y C(Y, Y ) + Y + θ XZ C(Z, Z) + P C(Z, Z) + R Z C(Z, Z) + Z + θ XY C(X, X ) + X C(Y, Y ) + Y C(Z, Z) + Z

Invariants and Tableaux Introduction We state invariants for Clifford gates based on the tableau of the gate. Tableau Intuition Clifford gates Linear transformations Tableaux Matrices

Invariants and Tableaux Introduction We state invariants for Clifford gates based on the tableau of the gate. Tableau Intuition Clifford gates Linear transformations Tableaux Matrices Matrix Representation Fix a basis for the vector space. Write the image of each basis element as a (linear) combination of basis elements. Format data as a table.

Invariants and Tableaux Introduction We state invariants for Clifford gates based on the tableau of the gate. Tableau Intuition Clifford gates Linear transformations Tableaux Matrices Matrix Tableau Representation Fix a basis for the vector space Pauli group. Write the image of each basis element as a (linear) combination of basis elements. Format data as a table.

Invariants and Tableaux Introduction We state invariants for Clifford gates based on the tableau of the gate. Tableau Intuition Clifford gates Linear transformations Tableaux Matrices Matrix Tableau Representation Fix a basis for the vector space Pauli group. Write the image of each basis element as a (linear) combination of basis elements. Format data as a table. Result: 2n 2n boolean matrix, length 2n bit vector (for phase).

Tableau Examples Introduction E.g., CSIGN and H have the following tableaux, ignoring phase bits. 1 0 0 1 ( CSIGN = 0 1 0 0 0 1 0 1 1 0 H = 1 0 0 0 0 1 )

Tableau Examples Introduction E.g., CSIGN and H have the following tableaux, ignoring phase bits. 1 0 0 1 ( CSIGN = 0 1 0 0 0 1 0 1 1 0 H = 1 0 0 0 0 1 ) CSIGN(X I ) CSIGN = (X I )(I Z). Basis is X I, Z I, I X, I Z. First row is 1, 0, 0, 1

Invariant Example Introduction CSIGN = 1 0 0 1 0 1 0 0 0 1 1 0 0 0 0 1 H = ( 0 1 1 0 ) Definition A gate is Z-preserving if there is a 0 in the bottom left of every block of the tableau. CSIGN is Z-preserving, H is not.

Outline Introduction Circuit Model Quick Tour of the Lattice Important gates Broad structural description Elements of the proof Universal construction

Circuit Model Section 2 Circuit Model

Axiom 1 Circuit Model Given circuits for U and V, we have circuits for UV and U V. Composition U V Tensor Product U V

Axiom 2 (Permutation) Circuit Model 1 Does locality matter? 2 Does order matter?

Axiom 2 (Permutation) Circuit Model 1 Does locality matter? 2 Does order matter? Our Solution Assume SWAP is in our gate set.

Axiom 3 (Ancillas) Circuit Model Our ancilla policy allows 1 arbitrarily many ancilla qubits,

Axiom 3 (Ancillas) Circuit Model Our ancilla policy allows 1 arbitrarily many ancilla qubits, 2 initialized to arbitrarily complex quantum states,

Axiom 3 (Ancillas) Circuit Model Our ancilla policy allows 1 arbitrarily many ancilla qubits, 2 initialized to arbitrarily complex quantum states, 3 but they must be returned to original state.

Ancilla Qubit Objections Circuit Model In principle, couldn t the ancilla states be impractical to construct?

Ancilla Qubit Objections Circuit Model In principle, couldn t the ancilla states be impractical to construct? Maybe, but they only need to be constructed once, then reused. For this classification we use a discrete set of 1 and 2 qubit ancilla states.

Ancilla Qubit Objections Circuit Model In principle, couldn t the ancilla states be impractical to construct? Maybe, but they only need to be constructed once, then reused. For this classification we use a discrete set of 1 and 2 qubit ancilla states. Clifford gates and magic states are universal for quantum computation, aren t they?

Ancilla Qubit Objections Circuit Model In principle, couldn t the ancilla states be impractical to construct? Maybe, but they only need to be constructed once, then reused. For this classification we use a discrete set of 1 and 2 qubit ancilla states. Clifford gates and magic states are universal for quantum computation, aren t they? Yes, but only if the magic states are destroyed in the process, and our ancillas must be returned intact.

Tour of the Lattice Section 3 Tour of the Lattice

Reminder Tour of the Lattice Theorem Given a collection G of Clifford gates, G is one of 57 possible classes. The 57 classes break into two major groups: 30 classes generated by single qubit gates, and 27 other classes.

Tour of the Lattice P + Γ θ +++ θ + θ + θ + P + R X P + R Y P + R Z P R X R Y R Z X + θ YZ Y + θ XZ Z + θ XY X Y Z Γ +++ Γ + Γ + Γ + θ Y +Z θ Y Z θ X +Z θ X Z θ X +Y θ X Y

Tour of the Lattice Understanding Single Qubit Classes Fact Each subgroup of C 1 defines a class. Each class generated by single qubit gates corresponds to a subgroup of C 1.

Tour of the Lattice Understanding Single Qubit Classes Fact Each subgroup of C 1 defines a class. Each class generated by single qubit gates corresponds to a subgroup of C 1. C 1 is isomorphic to the symmetries of the cube, or S 4. These have 30 subgroups, hence 30 classes.

Tour of the Lattice ALL C(Y, X ) + P + R X T4 + P + Γ C(Z, Y ) + P + R Y C(X, Z) + P + R Z C(X, Z) + P T4 + P + R X C(Y, X ) + P T4 + P + R Y C(Z, Y ) + P T4 + P + R Z T4 + P C(X, X ) + P + R X C(Y, Y ) + P + R Y C(Z, Z) + P + R Z C(X, X ) + P C(X, X ) + R X C(X, X ) + X + θ YZ C(Y, Y ) + P C(Y, Y ) + R Y C(Y, Y ) + Y + θ XZ C(Z, Z) + P C(Z, Z) + R Z C(Z, Z) + Z + θ XY C(X, X ) + X C(Y, Y ) + Y C(Z, Z) + Z

Color Legend Tour of the Lattice The colors indicate the gate is X -, Y -, or Z-preserving. Definition A gate is Z-preserving if there is a 0 in the bottom left of every block of the tableau. E.g., CSIGN, 1 0 0 1 0 1 0 0 0 1 1 0 0 0 0 1.

Preservation invariants Tour of the Lattice There are similar invariants for X - and Y -preserving gates. The invariants impose the following restrictions on a typical block ( a b c d ). X -preserving: b = 0, Y -preserving: a + b + c + d = 0, Z-preserving: c = 0. So CSIGN is not X -preserving or Y -preserving. 1 0 0 1 0 1 0 0 0 1 1 0 0 0 0 1.

Preservation invariants Tour of the Lattice There are similar invariants for X - and Y -preserving gates. The invariants impose the following restrictions on a typical block ( a b c d ). X -preserving: b = 0, Y -preserving: a + b + c + d = 0, Z-preserving: c = 0. But CNOT is X - and Z-preserving! 1 0 0 0 0 1 0 1 1 0 1 0 0 0 0 1.

Tour of the Lattice ALL C(Y, X ) + P + R X T4 + P + Γ C(Z, Y ) + P + R Y C(X, Z) + P + R Z C(X, Z) + P T4 + P + R X C(Y, X ) + P T4 + P + R Y C(Z, Y ) + P T4 + P + R Z T4 + P C(X, X ) + P + R X C(Y, Y ) + P + R Y C(Z, Z) + P + R Z C(X, X ) + P C(X, X ) + R X C(X, X ) + X + θ YZ C(Y, Y ) + P C(Y, Y ) + R Y C(Y, Y ) + Y + θ XZ C(Z, Z) + P C(Z, Z) + R Z C(Z, Z) + Z + θ XY C(X, X ) + X C(Y, Y ) + Y C(Z, Z) + Z

Tour of the Lattice ALL C(Y, X ) + P + R X T4 + P + Γ C(Z, Y ) + P + R Y C(X, Z) + P + R Z C(X, Z) + P T4 + P + R X C(Y, X ) + P T4 + P + R Y C(Z, Y ) + P T4 + P + R Z T4 + P C(X, X ) + P + R X C(Y, Y ) + P + R Y C(Z, Z) + P + R Z C(X, X ) + P C(X, X ) + R X C(X, X ) + X + θ YZ C(Y, Y ) + P C(Y, Y ) + R Y C(Y, Y ) + Y + θ XZ C(Z, Z) + P C(Z, Z) + R Z C(Z, Z) + Z + θ XY C(X, X ) + X C(Y, Y ) + Y C(Z, Z) + Z

Tour of the Lattice ALL C(Y, X ) + P + R X T4 + P + Γ C(Z, Y ) + P + R Y C(X, Z) + P + R Z C(X, Z) + P T4 + P + R X C(Y, X ) + P T4 + P + R Y C(Z, Y ) + P T4 + P + R Z T4 + P C(X, X ) + P + R X C(Y, Y ) + P + R Y C(Z, Z) + P + R Z C(X, X ) + P C(X, X ) + R X C(X, X ) + X + θ YZ C(Y, Y ) + P C(Y, Y ) + R Y C(Y, Y ) + Y + θ XZ C(Z, Z) + P C(Z, Z) + R Z C(Z, Z) + Z + θ XY C(X, X ) + X C(Y, Y ) + Y C(Z, Z) + Z

T 4 gate Tour of the Lattice Definition Let T 4 be the 4 qubit gate which does nothing to even-parity inputs, flips all bits of odd-parity inputs. E.g. T 4 0000 = 0000, and T 4 0001 = 1110.

T 4 tableau Tour of the Lattice The tableau for T 4 : Amazingly, T 4 is X -, Y -, and Z-preserving! 0 0 1 0 1 0 1 0 0 0 0 1 0 1 0 1 1 0 0 0 1 0 1 0 0 1 0 0 0 1 0 1 1 0 1 0 0 0 1 0 0 1 0 1 0 0 0 1 1 0 1 0 1 0 0 0 0 1 0 1 0 1 0 0

Symmetry? Tour of the Lattice ALL C(Y, X ) + P + R X T4 + P + Γ C(Z, Y ) + P + R Y C(X, Z) + P + R Z C(X, Z) + P T4 + P + R X C(Y, X ) + P T4 + P + R Y C(Z, Y ) + P T4 + P + R Z T4 + P C(X, X ) + P + R X C(Y, Y ) + P + R Y C(Z, Z) + P + R Z C(X, X ) + P C(X, X ) + R X C(X, X ) + X + θ YZ C(Y, Y ) + P C(Y, Y ) + R Y C(Y, Y ) + Y + θ XZ C(Z, Z) + P C(Z, Z) + R Z C(Z, Z) + Z + θ XY C(X, X ) + X C(Y, Y ) + Y C(Z, Z) + Z

Three-fold symmetry Tour of the Lattice The automorphism of the Pauli group that maps I I X Y Y Z Z X, lifts to an automorphism of the Clifford group, and an automorphism of the lattice.

Tour of the Lattice Z-preserving and misc. classes ALL T 4 + P + Γ CNOT +P + S CNOT +P T 4 + P + S T 4 + P CSIGN +P + S CSIGN +P CSIGN +S CSIGN +Z + θ XY CSIGN +Z

Proof Techniques Section 4 Proof Techniques

Proof Techniques Proof Techniques Steps of the proof:

Proof Techniques Proof Techniques Steps of the proof: 1 Characterize each class by generators and an invariant. E.g., T 4 generates all gates that are simultaneously X -, Y -, and Z-preserving.

Proof Techniques Proof Techniques Steps of the proof: 1 Characterize each class by generators and an invariant. E.g., T 4 generates all gates that are simultaneously X -, Y -, and Z-preserving.

Proof Techniques Proof Techniques Steps of the proof: 1 Characterize each class by generators and an invariant. E.g., T 4 generates all gates that are simultaneously X -, Y -, and Z-preserving. 2 Argue that the classes are distinct. E.g., CNOT is not in the class generated by T 4 because CNOT is not Y -preserving.

Proof Techniques Proof Techniques Steps of the proof: 1 Characterize each class by generators and an invariant. E.g., T 4 generates all gates that are simultaneously X -, Y -, and Z-preserving. 2 Argue that the classes are distinct. E.g., CNOT is not in the class generated by T 4 because CNOT is not Y -preserving.

Proof Techniques Proof Techniques Steps of the proof: 1 Characterize each class by generators and an invariant. E.g., T 4 generates all gates that are simultaneously X -, Y -, and Z-preserving. 2 Argue that the classes are distinct. E.g., CNOT is not in the class generated by T 4 because CNOT is not Y -preserving. 3 Argue that no class has been omitted.

No missing classes Proof Techniques 1 Suppose for a contradiction that G is not a listed class.

No missing classes Proof Techniques 1 Suppose for a contradiction that G is not a listed class. 2 For each invariant G does not satisfy, construct a simple gate violating that invariant. E.g., if some gate is not Y -preserving, extract one of a single qubit gate that is not Y -preserving, a CSIGN, a CNOT, etc.

No missing classes Proof Techniques 1 Suppose for a contradiction that G is not a listed class. 2 For each invariant G does not satisfy, construct a simple gate violating that invariant. E.g., if some gate is not Y -preserving, extract one of a single qubit gate that is not Y -preserving, a CSIGN, a CNOT, etc. 3 Use simple gates to construct canonical class generators. E.g., the class is defined as generated by T 4 and S, but you have T 4 and CSIGN.

Proof Techniques 1 Identify 2 2 block of the tableau violating the invariant.. 0 1 1 0. 2 Apply a SWAP to make that row the first one. 0 1 1 0.

Proof Techniques Assume the first row is ( a1 a 2... a 2k b 1... b l 0... 0 ) where a 1,..., a 2k are invertible 2 2 blocks, b 1,..., b l 0 are non-invertible 2 2 blocks.

Universal Construction Proof Techniques. U. U 1.

Proof Techniques b 1 b 2... b l b l... b 2 b 1 G(a 2 ) G 1 (a 2 ) G(a 3 ) G 1 (a 3 ). T 2k. G(a 2k ) G 1 (a 2k ) b 1 b 1 b 2 b 2...... b l b l

Open Problems Section 5 Open Problems

Open Problems Complete the classification of quantum gates. Possible strategy: divide into cases by group of single qubit gates.

Open Problems Complete the classification of quantum gates. Possible strategy: divide into cases by group of single qubit gates. Classify Clifford gates but with stabilizer state ancillas. We can show nothing changes unless Γ Γ 1 Γ, where Γ = HS.