Missingness Bias in Model Debugging
Abstract
Missingness, or the absence of features from an input, is a concept fundamental to many model debugging tools. However, in computer vision, pixels cannot simply be removed from an image. One thus tends to resort to heuristics such as blacking out pixels, which may in turn introduce bias into the debugging process. We study such biases and, in particular, show how transformer-based architectures can enable a more natural implementation of missingness, which side-steps these issues and improves the reliability of model debugging in practice.
Cite
Text
Jain et al. "Missingness Bias in Model Debugging." International Conference on Learning Representations, 2022.Markdown
[Jain et al. "Missingness Bias in Model Debugging." International Conference on Learning Representations, 2022.](https://mlanthology.org/iclr/2022/jain2022iclr-missingness/)BibTeX
@inproceedings{jain2022iclr-missingness,
title = {{Missingness Bias in Model Debugging}},
author = {Jain, Saachi and Salman, Hadi and Wong, Eric and Zhang, Pengchuan and Vineet, Vibhav and Vemprala, Sai and Madry, Aleksander},
booktitle = {International Conference on Learning Representations},
year = {2022},
url = {https://mlanthology.org/iclr/2022/jain2022iclr-missingness/}
}