Contributio of row/colum of a Hermitian matrix to the leading eigenvalue

Hi all, thanks in advance that if you can give any clues for my question below:

Assume that I have a Hermitian matrix (M = M'), matrix is square. If I do the singular value decomposition (SVD): [U,S,V] = svd(M), we can get the leading eigenvalue which is the 1st element in S. But how much contribution that each column/row is given to this leading eigenvalue? Is there a way to measure this kind of contribution (not the eigenvalue sensitivity)?