Grover's algorithm on a balanced function
Let \( n \) be a positive integer and let \( f: \Sigma^{n} \rightarrow \Sigma \) be a balanced function. Thus, for
\[
A_{0}=\left\{x \in \Sigma^{n}: f(x)=0\right\} \quad \text { and } \quad A_{1}=\left\{x \in \Sigma^{n}: f(x)=1\right\},
\]
it is the case that \( \left|A_{0}\right|=\left|A_{1}\right|=2^{n-1} \).
Which of the following statements best describes what happens when Grover's algorithm is run on the function \( f \) for some number of iterations, followed by a standard basis measurement of the \( n \) qubit register used in the algorithm?
Answer options
1. After a single iteration, the resulting output string \( x \) is guaranteed to satisfy \( f(x)=1 \), with probability 1 .
2. After two iterations, the resulting output string \( x \) is guaranteed to satisfy \( f(x)=1 \), with probability 1 .
3. After any number of iterations, the resulting output string \( x \) satisfies \( f(x)=1 \) with probability \( 1 / 2 \).
4. After any number of iterations, the resulting output string \( x \) never satisfies \( f(x)=1 \); it is certain to satisfy \( f(x)=0 \).