AI RESEARCH
Directional Consistency as a Complementary Optimization Signal: The GONO Framework
arXiv CS.LG
•
ArXi:2605.06575v1 Announce Type: new We identify and formalize an underexplored phenomenon in deep learning optimization: directional alignment and loss convergence can be decoupled. An optimizer can exhibit near-perfect directional consistency (cc_t -> 1, measured via consecutive gradient cosine similarity) while the loss remains high or decreases slowly.