Beyond Confidence: Reliable Models Should Also Consider Atypicality

TL;DR

This paper explores how considering the atypicality of inputs alongside confidence can enhance the reliability and accuracy of machine learning models, especially in uncertain or ambiguous cases.

Contribution

It introduces the concept of incorporating atypicality into uncertainty estimation, demonstrating its benefits for model calibration and performance across various models and tasks.

Findings

Atypical inputs are more overconfident and less accurate.

Incorporating atypicality improves uncertainty quantification.

Atypicality enhances model performance in real-world applications.

Abstract

While most machine learning models can provide confidence in their predictions, confidence is insufficient to understand a prediction's reliability. For instance, the model may have a low confidence prediction if the input is not well-represented in the training dataset or if the input is inherently ambiguous. In this work, we investigate the relationship between how atypical(rare) a sample or a class is and the reliability of a model's predictions. We first demonstrate that atypicality is strongly related to miscalibration and accuracy. In particular, we empirically show that predictions for atypical inputs or atypical classes are more overconfident and have lower accuracy. Using these insights, we show incorporating atypicality improves uncertainty quantification and model performance for discriminative neural networks and large language models. In a case study, we show that using atypicality improves the performance of a skin lesion classifier across different skin tone groups without having access to the group attributes. Overall, we propose that models should use not only confidence but also atypicality to improve uncertainty quantification and performance. Our results demonstrate that simple post-hoc atypicality estimators can provide significant value.