AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Cross-Task Knowledge Distillation in Multi-Task Recommendation

Chenxiao Yang, Junwei Pan, Xiaofeng Gao, Tingyu Jiang, Dapeng Liu, Guihai Chen

AAAI 2022 · Label-efficient and robust learning › Knowledge distillation · approach: Patient consistent distillation · 37 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add auxiliary fine-grained ranking relations (ablation)CrossDistil · TikTok TaskAMulti-AUC: 0.6510 -> 0.7411Improved
add auxiliary fine-grained ranking relations (ablation)CrossDistil · TikTok TaskBMulti-AUC: 0.8005 -> 0.8341Improved
calibrate teacher predictions before distillation (ablation)CrossDistil · TikTok TaskAMulti-AUC: 0.7396 -> 0.7411Improved
calibrate teacher predictions before distillation (ablation)CrossDistil · TikTok TaskBMulti-AUC: 0.8312 -> 0.8341Improved
apply teacher error correction (ablation)CrossDistil · TikTok TaskAMulti-AUC: 0.7399 -> 0.7411Improved
apply teacher error correction (ablation)CrossDistil · TikTok TaskBMulti-AUC: 0.8310 -> 0.8341Improved
transfer fine-grained knowledge from auxiliary ranking teachers rather than same-task regression teachers (ablation)CrossDistil · TikTok TaskAMulti-AUC: 0.6901 -> 0.7411Improved
transfer fine-grained knowledge from auxiliary ranking teachers rather than same-task regression teachers (ablation)CrossDistil · TikTok TaskBMulti-AUC: 0.8014 -> 0.8341Improved
avoid direct cross-task distillation by using conflict-filtering auxiliary ranking tasks (ablation)CrossDistil · TikTok TaskAAUC: 0.7269 -> 0.7494Improved
avoid direct cross-task distillation by using conflict-filtering auxiliary ranking tasks (ablation)CrossDistil · TikTok TaskBAUC: 0.9184 -> 0.9513Improved
use calibration (ablation)CrossDistil · TikTok TaskALogLoss: 0.5832 to 0.5703Improved
use calibration (ablation)CrossDistil · TikTok TaskBLogLoss: 0.0623 to 0.0337Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy