AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Instance-Dependent Early Stopping for Adaptive Data Pruning

Suqin Yuan, Runqi Lin, Felix Azian, Lei Feng, Bo Han, Gang Niu, Masashi Sugiyama, Tongliang Liu

TPAMI 2026 · Architectures and efficiency › RNNs and normalization · approach: Dataset distillation and data-efficient training · 5 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
stop backprop on mastered instances (IES)ResNet-18 training · CIFAR-10test accuracy: 92.9% -> 92.9% with 54.6% mini-batches savedNo clear effect
stop backprop on mastered instances (IES)SGD(F) ResNet-18 · CIFAR-10test accuracy: 92.1% -> 92.4%No clear effect
pretrain with IES before head fine-tuningResNet-101 · ImageNet-1k to CIFAR-10transfer accuracy: 80.3% -> 81.2%Improved
pretrain with IES before head fine-tuningResNet-101 · ImageNet-1k to CIFAR-100transfer accuracy: 55.6% -> 57.5%Improved
pretrain with IES before head fine-tuningResNet-101 · ImageNet-1k to Caltech-101transfer accuracy: 57.4% -> 59.9%Improved
use IES instead of conventional early stopping at ~2x speedupResNet · CIFAR-10test accuracy: 90.4% -> 93.7%Improved
use IES instead of conventional early stopping at ~3x speedupResNet · CIFAR-100test accuracy: 63.9% -> 73.0%Improved
use IES instead of EfficientTrain at ~2x speedupResNet · CIFAR-100test accuracy: 75.0% -> 74.9%No clear effect
use second-order loss difference criterion instead of loss value or other orders (ablation)IES · CIFAR-10 and CIFAR-100test accuracyImproved
select instances by IES rather than random removal (ablation)backprop sample selection · CIFAR-10 and CIFAR-100test accuracyImproved
vary delta over four orders of magnitude (ablation)IES · CIFARtest accuracy: within approximately 2% of baselineNo clear effect
apply IES under label noise (20% and 40%)IES · CIFAR-10 symmetric 20% noisebest accuracy: 87.81% -> 87.81%, 0% mini-batch savedNo clear effect

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