A remark on Khovanov homology and two-fold branched covers (Q968601): Difference between revisions

The author answers in the negative a question of Ozsváth, asking whether the total rank of the reduced Khovanov homology over the field with two elements is an invariant of \(2\)-fold branched covers of knots. The result is achieved by exhibiting two specific examples. These are \(2\)-fold branched covers of four specific torus knots (\(T(5,10n\pm 1)\), for \(n=1,2\)). It is easy to see that these manifolds are also \(2\)-fold branched covers of Montesinos knots. The conclusion follows by computing and comparing the ranks of the reduced Khovanov homologies for the two types of knots. Computations are obtained using the programme JavaKh by Bar-Natan and Green. The author concludes by observing that, conjecturally, there should be a discrepancy between the two ranks for all \(n>0\), however computations for large torus knots are too involved for this to be verified.