How to perform a sparse eigendecomposition in order to compute only the eigenvectors of smallest eigenvalues?

I have a large n * n sparse matrix called L. k is a given value. I am supposed to do this:

"perform a sparse eigendecomposition of the Laplacian L that computes only the eigenvectors associated with the k smallest-magnitude eigenvalues."

How can I do this?

This is what I have tried:

d = diag(eigs(L,k,'smallestabs'));

===============================================================

And then I am supposed to do this:

"Then project matrix V (vertex positions) onto the basis spanned by these eigenvectors and reconstruct a filtered V called V_new (version of the mesh)."

P is not important here. I just want the new V based on the D matrix. Basically I want to claculate this:

I have tried this:

pd = padarray(diag(d),[n-k,n-k],0,'post');
V_new = V * pd * V'

But seems not to be working, because the resulting V_new is very different than V.

How can I do this?