The NEIC moment tensor, if this is the quick one, is made rapidly with fewer stations and fewer waves, aiming to provide information for rapid response. It is a bit suspect in having a depth of 6km, too shallow for the centroid of an M6.8 event, and a very flat apparent dip, which could be right, but also is difficult to resolve.

The Harvard solution is an unusual mechanism - usually shallow faulting is either mostly thrust, strike-slip, or normal, but this is intermediate. It may well be a compound earthquake, breaking several fault planes with various orientations.