AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Pseudo-Non-Linear Data Augmentation: A Constrained Energy Minimization Viewpoint

Pingbang Hu, Mahito Sugiyama

ICLR 2026 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Disentangled and discrete latent models

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
augment training set with PNL samples (20% extra)MLP · Connectionist Benchtest accuracy (%): 88.10 -> 93.81Improved
augment training set with PNL samplesMLP · Wine Qualitytest accuracy (%): 55.00 -> 59.03Improved
augment training set with PNL samplesResNet-18 · CIFAR-10test accuracy (%): 88.57 -> 88.07Hurt
augment training set with PNL samplesMLP · MNISTtest accuracy (%): 97.98 -> 97.91No clear effect
augment training set with PNL samplesMLP · Taiwanese Bankruptcytest accuracy (%): 96.54 -> 96.53No clear effect
augment training set with PNL samplesM5 · Speech Commandstest accuracy (%): 84.48 -> 84.35No clear effect
use PNL instead of autoencoder augmentationMLP · Connectionist Benchtest accuracy (%): 82.86 -> 93.81Improved
use PNL instead of mixupMLP · MNISTtest accuracy (%): 96.45 -> 97.91Improved
PNL lowers variance across bootstrap testsMLP · Connectionist Benchaccuracy standard deviation: 8.58% -> 4.54%Improved
increase PNL augmentation size from 25% to 100% of original (ablation)MLP · MNISTtest accuracy (%): 98.01 -> 97.55Hurt
use k=1 nearest neighbor for the local data sub-manifold (ablation)PNL · MNIST (qualitative)sample diversityHurt
apply direct kernel density fitting in ambient space without dimension reduction (ablation)PNL · MNIST (1000 samples)augmentation qualityHurt

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