There isn't an easy way to answer that. A local added bit of lift (a bump) will be modeled as a horseshoe vortex, and the entire flowfield can be updated by adding the influence of this horseshoe vortex to the preceding flowfield. Thus a local bump in lift changes the entire wash, across the wing, outboard the wing, everywhere. If you go from positive lift to negative lift near the wingtip, you are adding a vortex with 'backwards' rotation, and the added vortex will increase downwash/decrease upwash across the entire span of the wing inboard of the 'bump' and outboard of the bump, but will have the opposite effect across the span of the bump itself. Pretend you are superimposing a similar, but smaller wing upside-down, and then add the similar, but weaker inverted flow field. None of this depends on whether the bump happens to be in an area upwash or downwash.
So if the reverse in lift occurs entirely outboard the vortex, the primary vortex moves outboard, perhaps right to the inboard bump vortex, then there is the outboard bump vortex. However, a smooth transition towards negative lift which begins inboard of the vortex would tend to move it inboard.
