Quantum Rozos
Quantum Rozos
Quantum Rozos
Marios Rozos
ECE-NTUA
June 2022
Vector Spaces: is a set with vectors as elements and closed under vector
addition and scalar multiplication.
3 Positive definiteness:
⟨x, x⟩ ≥ 0 if x ̸= 0
⟨x, x⟩ = 0 if x = 0
Outer Product
Given two vectors of size m × 1 and n × 1 respectively,
u = [u1 u2 . . . um ]T , v = [v1 v2 . . . vn ]T
Tensor product
Tensor product is defined by the following 3 properties:
c(|ψ1 ⟩ ⊗ |ψ2 ⟩) = (c |ψ1 ⟩) ⊗ |ψ2 ⟩ = |ψ1 ⟩ ⊗ (c |ψ2 ⟩)
(|ψ1 ⟩ + |φ1 ⟩) ⊗ |ψ2 ⟩ = |ψ1 ⟩ ⊗ |ψ2 ⟩ + |φ1 ⟩ ⊗ |ψ2 ⟩
|ψ1 ⟩ ⊗ (|ψ2 ⟩ + |φ2 ⟩) = |ψ1 ⟩ ⊗ |ψ2 ⟩ + |ψ1 ⟩ ⊗ |φ2 ⟩
for every c ∈ C, |ψ1 ⟩ , |φ1 ⟩ ∈ H1 , |ψ2 ⟩ , |ϕ2 ⟩ ∈ H2 .
15
Unitary operator
An operator U is called unitary if U † U = UU † = I , where I is the identity
operator.
The unitary operators preserve inner products between vectors and the
norms of them
Hermitean Operator
An operator T in a Hilbert space H is called Hermitean or self-adjoint if
T† = T
Projector
A projector on a vector space H is a linear operator P that satisfies
P 2 = P. An orthogonal projector is a projector that also satisfies P † = P.
T |ψ⟩ = c |ψ⟩
Postulate 1
Every isolated physical system can be associated with a Hilbert space
known as the state space of the system. The system is completely
described by its state vector, which is a unit vector in the system’s state
space.
Normalization Condition
|ψ⟩ is a unit vector, i.e. ⟨ψ|ψ⟩ = 1 or equivalently |α|2 + |β|2 = 1.
Superposition:
P A quantum state can be expressed as a linear combination
α
i i |ψi ⟩ of states |ψi ⟩ with coefficients (amplitudes) αi .
Example
The quantum state:
|0⟩ − |1⟩
√
2
√
is the√superposition of states |0⟩ and |1⟩ with amplitude 1/ 2 for |0⟩ and
−1/ 2 for the |1⟩.
We have:
1 0
|0⟩ = και |1⟩ =
0 1
and the set {|0⟩ , |1⟩} is called computational basis.
θ θ
|ψ⟩ = cos |0⟩ + e iφ sin |1⟩
2 2
Using θ and φ we can define a 3D-Sphere, called Block Sphere, which
helps us in qubit visualization.
We define:
1 1
|+⟩ ≡ √ |0⟩ + √ |1⟩
2 2
and
1 1
|−⟩ ≡ √ |0⟩ − √ |1⟩
2 2
|+⟩ and |−⟩ define Hadamard computational basis.
Postulate 2
The time-evolution of the state of a closed quantum system is described
by a unitary operator. That is, for any evolution of the closed system
there exists a unitary operator U such that if the initial state of the system
is |ψ1 ⟩, then after the evolution the state of the system will be
|ψ2 ⟩ = U |ψ1 ⟩
Postulate 3
When two physical systems are treated as one combined system, the
state space of the combined physical system is the tensor product
space H1 ⊗ H2 of the state spaces H1 , H2 of the component subsystems.
If the first system is in the state |ψ1 ⟩ and the second system in the state
|ψ2 ⟩, then the state of the combined system is:
|ψ1 ⟩ ⊗ |ψ2 ⟩
Postulate 4
The measurement of quantum states are described by a set {Mm } of
measurement operators, where m corresponds to one possible outcome.
The probability of measuring the mth state is
†
p(m) = ⟨ψ| Mm Mm |ψ⟩
and after the measurement the system collapses into the state
Mm |ψ⟩
†
⟨ψ| Mm Mm |ψ⟩
Measuring a qubit
Suppose we have a qubit |ψ⟩ = α |0⟩ + β |1⟩. Measurement operators are
M0 = |0⟩ ⟨0| and M1 = |1⟩ ⟨1|.
⇒ Quantum Mechanics
Pauli Gates
0 1 0 −i 1 0 1 0
X = , Y = , Z= , I =
1 0 i 0 0 −1 0 1
Hadamard Gate
1 1 1
H=√
2 1 −1
S Gate and T gate
1 0 1 0
S= , T =
0 i 0 e iπ/4
Pauli gates
X-gate: bit-shift (NOT gate)
Hadamard Gate
Hadamard gate turns a qubit into superposition of states (notice that
H2 = I )
1 1 1
H=√
2 1 −1
then:
1 1
H |0⟩ ≡ |+⟩ ≡ √ |0⟩ + √ |1⟩
2 2
and
1 1
H |1⟩ ≡ |−⟩ ≡ √ |0⟩ − √ |1⟩
2 2
Proof
Let’s say we have two slots, A and B. Slot A starts in an unknown but
pure quantum state |ψ⟩ and we want to copy it into slot B, while B starts
in some pure state |s⟩.
Proof (cont’d)
Suppose this copying procedure works for two particular pure states, |ψ⟩
and |ϕ⟩.
Then:
U(|ψ⟩ ⊗ |s⟩) = |ψ⟩ ⊗ |ψ⟩
U(|ϕ⟩ ⊗ |s⟩) = |ϕ⟩ ⊗ |ϕ⟩
Taking the inner product of these two equations gives
⟨ψ|ϕ⟩ = (⟨ψ|ϕ⟩)2
Proof (cont’d)
⟨ψ|ϕ⟩ = (⟨ψ|ϕ⟩)2
This means either |ψ⟩ = |ϕ⟩ or |ψ⟩ , |ϕ⟩ are orthogonal.
Thus a cloning device can only clone states which are orthogonal to one
another, and therefore a general quantum cloning device is impossible.
2
Nothing’s special about 3
The same holds if the algorithm gives the correct answer w.p. 1/2 + ϵ and
the wrong answer w.p. 1/2 − ϵ.
Since 1993 a central concern has been how BQP relates to classical
complexity classes, such as P, NP, and PH.
f (x) = x · s = x1 s1 ⊕ x2 s2 ⊕ · · · ⊕ xn sn ,
find s.
This problem was designed to probe an oracle separation between BQP
and BPP, i.e. there exists an oracle A s.t. BPPA ̸= BQPA .
Grover’s Algorithm
√
We can search any list of N items quantumly in O( N) queries.
P ⊆ NP ⊆ MA ⊆ QMA ⊆ PP ⊆ PSPACE
k-Local Hamiltonian
Given a k-local Hamiltonian on n qubits, ri=1 Hi , where r = poly(n) and
P
each Hi acts non-trivially on at most k qubits and has bounded operator
norm k ||H|| ≤ poly(n), determine whether:
(yes case) H has an eigenvalue less than a
(no case) all of the eigenvalues of H are larger than b, promised one
of these to be the case, where b − a = 1/poly (n)
NP ⊆ PostBQP
[Adleman, DeMorrai, Huang, ’91] PostBQP ⊆ PP
PostBQP is trivially closed under union, intersection, and complement
IP (Interactive Proof)
L ∈ IP:
x ∈ L ⇒ there exists prover P that the verifier accepts w.p. ≥ 2/3
x ̸∈ L ⇒ for all provers P the verifier rejects w.p. ≥ 2/3
Theorem (Shamir)
IP = PSPACE
Properties
QIP(1) is called QMA
QIP[k] = QIP[3] = QIP (Kitaev and Watrous)
QIP = IP = PSPACE (Jain, Ji, Upadhyay, and Watrous)
Marios Rozos (ECE-NTUA) Quantum Complexity Theory June 2022 68 / 74
QMIP, MIP*
Properties
NEXP ⊆ MIP*
QMIP = MIP*
MIP* = NEXP
4 There are types of correlations between Alice and Bob that can be
produced using infinite entanglement, but that can’t even be
approximated using any finite amount of entanglement.
5 The undecidability result disproves the Connes embedding conjecture,
a central conjecture from the theory of operator algebras and the
Tsirelson conjecture in quantum information theory.
6 It is one of the first non-relativizing results in quantum computability
theory.