Thompson's formula says, given $A,B \in \mathfrak{su}(n)$, there exists $U,V \in SU(n)$ such that:

$e^{A}e^{B}=e^{UAU^{\dagger} + VBV^{\dagger}}$

Given $a,b \in \mathfrak{su}(4)$ defined by: $a=J_x i\sigma_x\otimes\sigma_x + J_y i\sigma_x\otimes\sigma_y + J_z i\sigma_z\otimes\sigma_z$ for all $J_k$ real and non-zero, $\sigma_k$ the Pauli matrices and $b=i\sigma_z\otimes I$, consider the function $F:\mathbb{R}^2\rightarrow SU(4)$ defined by:


Is it actually possible to find some suitable $U,V$ in this case in terms of $a,b$ and $w_1,w_2$ and thus write $F$ explicitly in the form of the RHS of Thompson's formula? If so, how can I do this without simply trying to explicitly evaluate the exponentials.

  • 2
    $\begingroup$ I assume that you know that $U$ and $V$ are very far from being unique in Thompson's formula. In fact, for 'generic' $A$ and $B$, there will be a family of pairs $(U,V)$ of dimension $n^2-1$ that satisfy Thompson's formula. Just look at the case $n=2$, and you'l see why. $\endgroup$ – Robert Bryant Sep 14 '15 at 16:23
  • $\begingroup$ Yes, thanks for pointing that out. I did know and I've edited accordingly. Finding a family of them would be extra interesting... $\endgroup$ – Benjamin Sep 14 '15 at 17:07

Your Answer

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

Browse other questions tagged or ask your own question.