On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,695 papers · 148 categories

Trend · papers per month

count % of corpus

2 results for “auto-augmentation”

Improving Auto-Augment via Augmentation-Wise Weight Sharingcs.LG

This paper improves auto-augment efficiency by sharing augmentation weights.

problem Efficient evaluation of augmentation policies for model training.

method Augmentation-Wise Weight Sharing (AWS) to create a fast yet accurate proxy task.

result Augmentation policies found achieve superior accuracies compared to existing methods.

auto-augmentation augmentation-wise weight sharing proxy task efficient evaluation model training

2020-09-30abs ↗pdf ↗

Drop-Activation: Implicit Parameter Reduction and Harmonic Regularizationcs.LG

Overfitting frequently occurs in deep learning. In this paper, we propose a novel regularization method called Drop-Activation to reduce overfitting and improve generalization. The key idea is to drop nonlinear activation functions by setting them to be identity functions randomly during training time. During testing, …

2018-11-14abs ↗pdf ↗