AI RESEARCH
Divergence-Suppressing Couplings for Rectified Flow
arXiv CS.LG
•
ArXi:2605.17733v1 Announce Type: cross The promise of Rectified Flow rests on producing self-generated couplings whose trajectories are straight, or nearly so. In practice, trajectories generated by the base flow model can bend and intertwine, and the resulting coupling inherits this distortion. In this paper, we identify that such trajectory entanglement is often associated with regions of nonzero divergence in the learned velocity field, where local expansion or contraction distorts trajectories and steers particles away from their ideal endpoints.