Identifying Spurious Biases Early in Training through the Lens of Simplicity Bias

by   Yu Yang, et al.

Neural networks trained with (stochastic) gradient descent have an inductive bias towards learning simpler solutions. This makes them highly prone to learning simple spurious features that are highly correlated with a label instead of the predictive but more complex core features. In this work, we show that, interestingly, the simplicity bias of gradient descent can be leveraged to identify spurious correlations, early in training. First, we prove on a two-layer neural network, that groups of examples with high spurious correlation are separable based on the model's output, in the initial training iterations. We further show that if spurious features have a small enough noise-to-signal ratio, the network's output on the majority of examples in a class will be almost exclusively determined by the spurious features and will be nearly invariant to the core feature. Finally, we propose SPARE, which separates large groups with spurious correlations early in training, and utilizes importance sampling to alleviate the spurious correlation, by balancing the group sizes. We show that SPARE achieves up to 5.6 worst-group accuracy than state-of-the-art methods, while being up to 12x faster. We also show the applicability of SPARE to discover and mitigate spurious correlations in Restricted ImageNet.


page 4

page 8

page 9

page 24

page 25


Is Last Layer Re-Training Truly Sufficient for Robustness to Spurious Correlations?

Models trained with empirical risk minimization (ERM) are known to learn...

Neural networks trained with SGD learn distributions of increasing complexity

The ability of deep neural networks to generalise well even when they in...

Learning an Invertible Output Mapping Can Mitigate Simplicity Bias in Neural Networks

Deep Neural Networks are known to be brittle to even minor distribution ...

An Investigation of Why Overparameterization Exacerbates Spurious Correlations

We study why overparameterization – increasing model size well beyond th...

Which Features are Learnt by Contrastive Learning? On the Role of Simplicity Bias in Class Collapse and Feature Suppression

Contrastive learning (CL) has emerged as a powerful technique for repres...

Robust Learning with Progressive Data Expansion Against Spurious Correlation

While deep learning models have shown remarkable performance in various ...

An adversarial feature learning strategy for debiasing neural networks

Simplicity bias is the concerning tendency of deep networks to over-depe...

Please sign up or login with your details

Forgot password? Click here to reset