Reflection: What You Can Now Build

You have worked through fifteen modules. Before moving to the next level it is worth pausing to take stock of every tool you now carry — and to notice how much territory those tools already cover.

The single-qubit layer

A single qubit lives in a two-dimensional complex vector space. Its state is a unit vector

which you can visualize as a point on the Bloch sphere: the north pole is $|0\rangle$ , the south pole is $|1\rangle$ , and the equator is the belt of equal-weight superpositions. Every single-qubit gate is a rotation of that sphere.

The gates you can now apply fall into a few families:

| Family | Gates | What they do | |---|---|---| | Pauli flips | $X$ , $Y$ , $Z$ | $180°$ rotations about $x$ , $y$ , $z$ | | Hadamard | $H$ | Swaps $z$ - and $x$ -axes; creates superposition | | Phase gates | $S$ , $S^\dagger$ , $T$ , $T^\dagger$ , $P(\lambda)$ | Rotations about $z$ by fixed or arbitrary angles | | Rotation gates | $R_x(\theta)$ , $R_y(\theta)$ , $R_z(\theta)$ | Continuous rotations about each axis |

Together, $H$ and $T$ are universal for single-qubit unitaries: $H$ and $T$ generate a dense subgroup of $\mathrm{SU}(2)$ , so any single-qubit operation can be approximated to any desired precision using only those two gates (Nielsen & Chuang, §4.2). The Solovay–Kitaev theorem further guarantees that the number of $H$ / $T$ gates needed scales only polylogarithmically in $1/\epsilon$ for precision $\epsilon$ . The rotation gates give you the same power with direct, continuous control.

The two-qubit layer

Two qubits live in a four-dimensional space spanned by $|00\rangle$ , $|01\rangle$ , $|10\rangle$ , $|11\rangle$ . The feature that has no classical analogue is entanglement: there exist two-qubit states — such as the Bell states — that cannot be written as a product of two independent single-qubit states.

The two-qubit gates you have used:

CNOT / CX — the workhorse of entanglement creation. Flips the target when the control is $|1\rangle$ .
CZ — applies a $\pi$ phase when both qubits are $|1\rangle$ ; symmetric in control and target.
CY, CP( $\lambda$ ) — conditional $Y$ and phase; same idea, different action.
SWAP — exchanges the states of two qubits in three CNOTs.
Toffoli (CCX) — a three-qubit gate; flips the target only when both controls are $|1\rangle$ . It makes the gate set classically universal as well.

Adding any entangling two-qubit gate to a universal single-qubit set yields a universal gate set for quantum computing — capable, in principle, of implementing any unitary on any number of qubits.

The measurement layer

Measurement in the computational basis produces a classical bit: outcome $b \in \{0, 1\}$ appears with probability $|\langle b|\psi\rangle|^2$ (the Born rule). After the measurement the qubit collapses to $|b\rangle$ . You can measure individual qubits or all qubits at once, and you have seen how to change the basis before measuring to extract information along a different axis — the trick behind state tomography.

What you have built in this module

The projects in Module 15 used all three layers together:

Quantum coin flip and RNG — Hadamard plus measurement, exploiting true randomness backed by quantum mechanics rather than a pseudo-random algorithm.
Biased coin — a $R_y(\theta)$ rotation before measurement, turning angle into probability via the Born rule: $P(1) = \sin^2(\theta/2)$ .
Bell pair and Bell test — CNOT after Hadamard creates maximal entanglement; changing the measurement basis reveals correlations that no classical model can reproduce.
Hidden-gate identification — probing an unknown unitary with carefully chosen inputs shows how to learn about a gate from its action on states.
State tomography — measuring in three mutually unbiased bases ( $Z$ , $X$ , $Y$ ) lets you reconstruct all three Bloch-vector components $\langle X\rangle$ , $\langle Y\rangle$ , $\langle Z\rangle$ .
Reusable circuit library — packaging state-preparation routines so they can be composed into larger circuits without copy-pasting.

Where this leads

The beginner track has equipped you with the language of quantum circuits. The intermediate track builds algorithms out of this language: Deutsch–Jozsa, Bernstein–Vazirani, Simon's algorithm, and Grover's search will each show you a setting where quantum interference lets a circuit find an answer in fewer queries than any classical algorithm can. Every one of those algorithms uses exactly the gates and measurement ideas you already know.