The state-of-the-art feature attribution methods often neglect the influence of unobservable confounders, posing a risk of misinterpretation, especially when it is crucial for the interpretation to remain faithful to the data. To counteract this, we propose a new approach, data-faithful feature attribution, which trains a confounder-free model using instrumental variables.

We investigate here the extent to which large language models pretrained on source code can solve PBE.

This occurs even under stationary data distributions, and its underlying mechanism is poorly understood.