1
$\begingroup$

Let's say we have the two many-body states $$ |\psi_k\rangle=\left(\prod_{k=1}^n c_k^\dagger\right)|0\rangle ,\qquad |\psi_\lambda\rangle=\left(\prod_{\lambda=1}^n c_\lambda^\dagger\right)|0\rangle \tag{1}$$ where $k$ and $\lambda$ label the states in different basis. These are related by $$ c_k=\sum_\lambda \langle k| \lambda \rangle c_\lambda ,\qquad c_k^\dagger=\sum_\lambda \langle \lambda |k\rangle c_\lambda^\dagger\tag{2}$$ Are there techniques to evaluate $\langle\psi_k|\psi_\lambda\rangle$? My first approach is to write

$$ |\psi_\lambda \rangle = \left(\prod_{\lambda'} \sum_k \langle \lambda'|k\rangle c_k^\dagger\right)|0\rangle\tag{3}$$ and then try to expand the sums and match operators in the same basis. This, however, seems to be a nightmare! Is there a different approach?

$\endgroup$
3
  • 2
    $\begingroup$ Now expand out (3) and collect all the coefficients corresponding to the state $|\psi_\lambda\rangle$. The result is a Slater determinant. $\endgroup$
    – Meng Cheng
    Mar 17, 2022 at 23:44
  • $\begingroup$ It looks like following this approach I will get that the answer is the determinant of the matrix $\langle k | \lambda \rangle$. Is this correct? $\endgroup$
    – Ivan
    Mar 17, 2022 at 23:49
  • $\begingroup$ that is correct. $\endgroup$
    – Meng Cheng
    Mar 18, 2022 at 0:01

1 Answer 1

1
$\begingroup$

Computing the overlap of many-body wavefunctions is in general a fairly difficult task. There are some clever algorithms that reduce the nightmarish-ness - you may find e.g. the one presented in this paper to be of interest.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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