Section 2.5 (4.3) of the Qiskit textbook, see here, discusses the conjugation of $R_x(\theta)$ by $CNOT$. The following expression is given:

$$CX_{j,k}(R_x(\theta)\otimes 1) CX_{j,k}=\color{brown}{CX_{j,k}e^{i\frac{\theta}{2}(X\otimes1)}CX_{j,k}=e^{i\frac{\theta}{2}CX_{j,k}(X\otimes 1)CX_{j,k}}}=e^{i\frac{\theta}{2}X\otimes X}$$

I am confused by the part highlighted in yellow. What exponentiation rules allow this? Could someone show me the intermediate steps or the relevant identities/rules to take us from the LHS to RHS of the highlighted part?


This is an application of the following identity

$$ Be^AB^{-1} = e^{BAB^{-1}}\tag1 $$

where $A$ is any $n\times n$ real or complex matrix and $B$ is any invertible $n\times n$ real or complex matrix.

Proof of $(1)$. First, recall that the matrix exponential of $A$ is defined as

$$ e^A = \sum_{k=0}^\infty \frac{1}{k!}A^k. $$

Next, note that for any integer $k$

$$ (BAB^{-1})^k = BAB^{-1}BAB^{-1}\dots BAB^{-1} = BAA\dots AB^{-1} = BA^kB^{-1}\tag2. $$

Finally, calculate

$$ e^{BAB^{-1}} = \sum_{k=0}^\infty \frac{1}{k!}(BAB^{-1})^k = \sum_{k=0}^\infty \frac{1}{k!}BA^kB^{-1} = B \left(\sum_{k=0}^\infty \frac{1}{k!}A^k\right) B^{-1} = Be^AB^{-1} $$

where we used $(2)$ in the second step. $\square$

  • $\begingroup$ Thanks so much, Adam! It now makes sense. $\endgroup$
    – Othman
    Jan 10 at 21:44
  • $\begingroup$ You're welcome! :-) $\endgroup$ Jan 10 at 21:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.