Google Research

Estimating Example Difficulty using Variance of Gradients (2020)


In machine learning, a question of great interest is understanding what examples are challenging for a model to classify. Identifying atypical examples helps inform safe deployment of models, isolates examples that require further human inspection, and provides interpretability into model behavior. In this work, we propose Variance of Gradients (VOG) as a proxy metric for detecting outliers in the data distribution. We provide quantitative and qualitative support that VOG is a meaningful way to rank data by difficulty and to surface a tractable subset of the most challenging examples for human-in-the-loop auditing. Data points with high VOG scores are more difficult for the model to classify and over-index on examples that require memorization.

Learn more about how we do research

We maintain a portfolio of research projects, providing individuals and teams the freedom to emphasize specific types of work