CliffRank: A Dual-Branch Framework for Activity-Cliff Ranking Prediction
CliffRank trains two parallel predictors with mean squared error, a thresholded listwise loss, and Pairwise Preference Consistency (PPC), which aligns relative ordering in the preference-probability space, and defines its practical limits.
Ke-Wei Li, Rong Zhang, Pei-Yu Yang et al.
· 0 citations