Example Perplexity

03/16/2022

∙

Some examples are easier for humans to classify than others. The same should be true for deep neural networks (DNNs). We use the term example perplexity to refer to the level of difficulty of classifying an example. In this paper, we propose a method to measure the perplexity of an example and investigate what factors contribute to high example perplexity. The related codes and resources are available at https://github.com/vaynexie/Example-Perplexity.

READ FULL TEXT

Example Perplexity

Visualizing Classification Structure in Deep Neural Networks

Synergy between Observation Systems Oceanic in Turbulent Regions

Understanding Rare Spurious Correlations in Neural Networks

Lazy vs hasty: linearization in deep networks impacts learning schedule based on example difficulty

Datamodels: Predicting Predictions from Training Data

Time Regularization in Optimal Time Variable Learning

LightGlue: Local Feature Matching at Light Speed

Example Perplexity

Related Research

Visualizing Classification Structure in Deep Neural Networks

Synergy between Observation Systems Oceanic in Turbulent Regions

Understanding Rare Spurious Correlations in Neural Networks

Lazy vs hasty: linearization in deep networks impacts learning schedule based on example difficulty

Datamodels: Predicting Predictions from Training Data

Time Regularization in Optimal Time Variable Learning

LightGlue: Local Feature Matching at Light Speed