AI Paper Atlas

Research map / Optimization / Optimizers

Module-Aware Optimization for Auxiliary Learning

Hong Chen, Xin Wang, Yue Liu, Yuwei Zhou, Chaoyu Guan, Wenwu Zhu

NeurIPS 2022 · Optimization › Optimizers · approach: Gradient-based meta-learning theory · 1 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
module-level auxiliary importance instead of single loss weights (Block-L vs Model-L) (ablation)MAOAL · Petaccuracy: 67.18 -> 69.61Improved
module-level importance instead of whole-model weighting (Block-L vs Model-L) (ablation)MAOAL · CUB-semiaccuracy: 51.16 -> 52.36Improved
finer partition: each parameter as a module instead of ResNet blocks (ablation)MAOAL · CUB-semiaccuracy: 52.36 -> 52.48No clear effect
finer partition: each parameter as a module instead of ResNet blocks (ablation)MAOAL · Petaccuracy: 69.61 -> 66.14Hurt
add module-aware auxiliary rotation loss versus single-loss trainingMAOAL vs SLL · Pet-semi (20% labels)accuracy: 31.81 -> 51.42Improved
add MAOAL auxiliary learning versus single-loss trainingMAOAL vs SLL · CUB-semi (20% labels)accuracy: 48.15 -> 52.36Improved
add MAOAL with harmful auxiliary losses versus single-loss trainingMAOAL vs SLL · CUBaccuracy: 77.29 -> 77.30No clear effect
equal-weight auxiliary losses with harmful auxiliaries versus single-loss trainingEqual weights · CUBaccuracy: 77.29 -> 71.68Hurt
MAOAL versus single-loss training with L2 regularizer as auxiliary on sparse dataMAOAL vs SLL · Amazon BeautyRMSE: 1.1008 -> 1.0873Improved
MAOAL versus single-loss training with L2 auxiliary on dense dataMAOAL vs SLL · MovieLens-1MRMSE: 0.9068 -> 0.9045No clear effect
rotation prediction auxiliary loss on low-resolution CIFAR imagesMAOAL vs SLL · CIFAR100accuracy: 46.52 -> 47.57Improved
more auxiliary tasks (NYUv2 depth and normal) with MAOAL versus single-lossMAOAL vs SLL · NYUv2 segmentationm-IOU: 32.61 -> 33.88Improved

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