DADF adds a plug-in second-stage debiasing layer with dynamic target transformation, duration-aware residual modeling, and multi-label auxiliary signals to reduce local calibration errors in long-tailed watch-time regression.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.IR 2years
2026 2representative citing papers
IEFF enables retrain-free feature efficiency rollouts in ranking systems by elastically controlling feature coverage at serving time, achieving 5x faster rollouts, zero retraining GPU cost, and 50-55% less performance degradation than abrupt feature removal.
citing papers explorer
-
DADF: A Distribution-Aware Debiasing Framework for Watch-Time Regression in Recommender Systems
DADF adds a plug-in second-stage debiasing layer with dynamic target transformation, duration-aware residual modeling, and multi-label auxiliary signals to reduce local calibration errors in long-tailed watch-time regression.
-
Intelligent Elastic Feature Fading: Enabling Model Retrain-Free Feature Efficiency Rollouts at Scale
IEFF enables retrain-free feature efficiency rollouts in ranking systems by elastically controlling feature coverage at serving time, achieving 5x faster rollouts, zero retraining GPU cost, and 50-55% less performance degradation than abrupt feature removal.