L3: Measurement, Bra Ket, Eigenvector, Eigenvalues
Overview
Hiu-Yung Wong develops the mathematics of quantum states from ket and bra notation through measurement, normalization, basis changes, and matrix operators. The lecture connects measurement probabilities to squared coefficient magnitudes, introduces eigenvectors and eigenvalues through the Pauli matrices, and works through how Pauli Z measures the computational basis while Pauli X flips |0⟩ and |1⟩.
Key takeaways
- A measurement of a state expressed as Σᵢ aᵢ|i⟩ returns basis state |i⟩ with probability |aᵢ|², and the amplitudes must be normalized so those probabilities sum to one.
- Taking the dual of c|a⟩ produces c*⟨a|, and reversing an inner product gives ⟨a|b⟩ = ⟨b|a⟩*.
- The same state can have different coordinate vectors in different bases: the equal superposition of |0⟩ and |1⟩ is |+⟩, certain to produce + in the plus-minus basis but giving either computational-basis outcome with 50 percent probability.
- A nonzero state |α⟩ is normalized by dividing by √⟨α|α⟩, making its squared amplitudes valid measurement probabilities.
- An eigenvector of operator A satisfies A|a⟩ = λ|a⟩; for Pauli Z, |0⟩ and |1⟩ have eigenvalues +1 and −1 respectively.
- Pauli X exchanges |0⟩ and |1⟩, so it acts as the quantum NOT gate on a single qubit.
Chapters
- A ket such as |a⟩ represents a vector by naming the state, while its coefficients specify components along a chosen basis.
- An n-dimensional basis labeled from 0 to n−1 has n basis vectors and can represent a state as a column of n coefficients.
- The same vector can be written using different basis labels, such as |0⟩ and |1⟩ or alternative names for the basis vectors.
- Measuring a superposition causes it to collapse to one of the basis states, with the probability of outcome i given by |aᵢ|².
- For a complex amplitude, the squared magnitude is aᵢaᵢ*, where aᵢ* is its complex conjugate.
- Hiu-Yung Wong emphasizes that measurement reveals one outcome, not every component of a superposition; quantum algorithms use interference to make desired outcomes more likely.
- The bra ⟨a| is the dual of ket |a⟩, formed by taking the transpose and complex conjugate of its column-vector representation.
- Multiplying a ket by a scalar c changes its bra to c*⟨a|, so the scalar is complex-conjugated under the dual operation.
- For a linear combination cα|a⟩ + cβ|b⟩, the corresponding bra is cα*⟨a| + cβ*⟨b|.
- The inner products satisfy ⟨a|b⟩ = ⟨b|a⟩*, rather than generally being equal as complex numbers.
- When an inner product is real, as in ordinary real-vector examples, complex conjugation leaves its value unchanged.
- A worked example with a vector containing i and a scalar c = 1 + i checks that taking the bra conjugates the scalar consistently.
- Two vectors are orthogonal when their inner product is zero, matching perpendicular vectors in three-dimensional geometry.
- A normalized vector has ⟨α|α⟩ = 1, meaning its length is one.
- For a quantum state with measurement amplitudes aᵢ, normalization ensures that the probabilities |aᵢ|² sum to one.
- A nonzero vector |α⟩ is normalized by dividing it by its length: |α⟩/√⟨α|α⟩.
- The denominator √⟨α|α⟩ is a real scalar, so the result remains a vector.
- Taking the inner product of the normalized vector with itself gives one, verifying the normalization procedure.
- A two-dimensional example represents the same vector using the |0⟩, |1⟩ basis and a basis rotated by 45 degrees, labeled |+⟩ and |−⟩.
- The state with equal amplitudes for |0⟩ and |1⟩ is the |+⟩ basis state.
- The column of coordinates changes with the basis, even though the underlying physical vector does not.
- Measuring the equal superposition of |0⟩ and |1⟩ in the |+⟩, |−⟩ basis yields |+⟩ with certainty.
- Measuring that same state in the computational basis gives |0⟩ and |1⟩ with equal, 50 percent probabilities.
- The example offers a qualitative analogy for position and momentum uncertainty, while the lecture notes that a full derivation is not covered.
- For a 45-degree basis rotation, |0⟩ is expressed as an equal combination of |+⟩ and |−⟩.
- The lecture converts a state by substituting the new-basis expressions for |0⟩ and |1⟩, then collecting the |+⟩ and |−⟩ coefficients.
- This coordinate conversion is compared to changing units, such as converting a dozen eggs or exchanging one currency for another.
- A quantum state is represented by a vector, and an operator is represented by a matrix acting on that vector.
- Matrix action generalizes the familiar geometric idea of rotating a vector to state spaces that cannot be visualized directly.
- Because physical quantum states must remain normalized, arbitrary stretching or shrinking is not a valid state evolution.
- A vector |a⟩ is an eigenvector of operator A when A|a⟩ = λ|a⟩ for a scalar eigenvalue λ.
- The eigenvalue describes how the operator acts on that particular vector without changing its direction, apart from the scalar factor.
- Eigenvalues are invariant under a change of basis, making them characteristic properties of an operator rather than coordinates tied to one representation.
- A spin-1/2 system has two basis states, conventionally called spin up and spin down or |0⟩ and |1⟩.
- The Pauli Z operator is the 2 × 2 diagonal matrix with entries +1 and −1.
- The lecture introduces Pauli matrices as central two-state quantum operators used repeatedly in quantum computing.
- Applying Pauli Z to |0⟩ returns |0⟩, so |0⟩ is an eigenvector with eigenvalue +1.
- Applying Pauli Z to |1⟩ returns −|1⟩, so |1⟩ is an eigenvector with eigenvalue −1.
- The diagonal entries of Pauli Z are its eigenvalues in the computational basis, and those eigenvalues remain unchanged under a basis transformation.
- Pauli X is the matrix with off-diagonal entries 1 and diagonal entries 0.
- Matrix multiplication shows that Pauli X maps |0⟩ to |1⟩, corresponding to a spin-up to spin-down flip.
- In quantum computing, Pauli X performs the NOT operation by exchanging the two computational-basis states.
Transcript excerpt
0:07 If not, uh, please study ASAP. It is not difficult for this class, but you need to keep track of it. Do not wait until midterm to start studying. That will be too late. You won't be able to because a lot of the concept, a lot intuitive. You need a lot of time, a lot of thinking throughout the semester to understand it. If you can do that, everything should be easy. So uh this week please finish chapter 1 to five again should be easy reading. So in the last lecture we talk about bracket notation right. So I just want to remind you you know this symbol uh a arrow which is a vector that is what we have been using in other classes and then we can represent it as in a cat
0:54 form. It is just a symbol also anything inside is to express what this head or this vector is about. Once we identify the basis then we can just write the components. So this is a z a1 all the way to a n. So for example right it it is n this is n + one dimensional in this case right. So it might have uh it might mean it might mean uh a equals to a z we z plus a1 a1 we hat right and then plus all the way to a
1:46 a uh n v n what right? So what how many dimension do we have in this space? n + one value because I start from zero just accidentally right n + one d right and again these are the basis vector and so that I just write them in the column form okay and same for b and then the definition of inner products turn out to
This is the start of the transcript. Save this video free for the whole transcript, searchable and time-coded.
In this course
Related notes
Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, Quantum Computing, TCAD, Semicond by Hiu-Yung Wong.