Bayes-TrEx: a Bayesian Sampling Approach to Model Transparency by Example
This provides a tool for researchers and practitioners to debug and understand model behaviors more flexibly than relying solely on test sets, though it is incremental as it builds on existing post-hoc explanation methods.
The paper tackles the challenge of interpreting neural networks by introducing Bayes-TrEx, a Bayesian sampling framework that finds in-distribution examples with specified prediction confidences, enabling analysis of high-confidence failures and ambiguous classifications across datasets like CLEVR, MNIST, and Fashion-MNIST.
Post-hoc explanation methods are gaining popularity for interpreting, understanding, and debugging neural networks. Most analyses using such methods explain decisions in response to inputs drawn from the test set. However, the test set may have few examples that trigger some model behaviors, such as high-confidence failures or ambiguous classifications. To address these challenges, we introduce a flexible model inspection framework: Bayes-TrEx. Given a data distribution, Bayes-TrEx finds in-distribution examples with a specified prediction confidence. We demonstrate several use cases of Bayes-TrEx, including revealing highly confident (mis)classifications, visualizing class boundaries via ambiguous examples, understanding novel-class extrapolation behavior, and exposing neural network overconfidence. We use Bayes-TrEx to study classifiers trained on CLEVR, MNIST, and Fashion-MNIST, and we show that this framework enables more flexible holistic model analysis than just inspecting the test set. Code is available at https://github.com/serenabooth/Bayes-TrEx.