AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Diversity-aware Weight Perturbation Promotes Robust Adaptation

Zibo Chen, Ruxin Li, Zilu Wang

ICML 2026 · Architectures and efficiency › RNNs and normalization · approach: Distributed LLM inference systems

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 diversity-aware affinity weighting on top of plain weight noise injection (ablation)DWP (ResNet) · CIFAR10 (mismatch 0.7)accuracy: additional 14.0% accuracy boostImproved
add plain multiplicative weight-noise injection during training (ablation)baseline ResNet · CIFAR10 (mismatch 0.7)accuracy: 12.0% to 50.0%Improved
increase noise realizations n from 5 to 10 (ablation)DWP (ResNet) · CIFAR10accuracy: 60.01 -> 65.61Improved
increase noise realizations n from 10 to 20 (ablation)DWP (ResNet) · CIFAR10accuracy: 65.61 -> 66.43; training time 32.32s -> 60.46sNo clear effect
match training mismatch level to test level (0.3 vs 0.1) (ablation)DWP (ResNet) · CIFAR10 (test mismatch 0.3)accuracy: 70.33 -> 87.54Improved
raise training mismatch level beyond test level (0.3 to 0.5) (ablation)DWP (ResNet) · CIFAR10 (test mismatch 0.3)accuracy: 87.54 -> 75.59Hurt
train with DWP instead of standard trainingF-MNIST CNN · F-MNIST (mismatch 0.7)accuracy: 53.95 -> 90.01Improved
train with DWP instead of standard trainingCIFAR10 ResNet · CIFAR10 (mismatch 0.5)accuracy: 18.02 -> 73.00Improved
train with DWP instead of standard trainingCIFAR10 ViT · CIFAR10 (mismatch 0.7)accuracy: 20.15 -> 65.30Improved
train with DWP instead of Forward Noise at low mismatchF-MNIST CNN · F-MNIST (mismatch 0.1)accuracy: 91.20 vs 91.74No clear effect
apply DWP to Qwen3-1.7B classification (clean accuracy cost)Qwen3-1.7B · iFLYTEK 119 (noise 0.2)clean accuracy: 58.06 -> 53.56Hurt
apply DWP to Qwen3-1.7B under strong noiseQwen3-1.7B · iFLYTEK 119 (noise 0.4)noisy accuracy: 1.26 -> 8.93Improved

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