Skip to content

README.md

Latest commit

 

History

History
62 lines (45 loc) · 3.37 KB

README.md

File metadata and controls

62 lines (45 loc) · 3.37 KB

Missingness Bias in Model Debugging.

This repository contains the code of our ICLR 2022 paper.

Missingness Bias in Model Debugging
Saachi Jain*, Hadi Salman*, Eric Wong, Pengchuan Zhang, Vibhab Vineet, Sai Vemprala, Aleksander Madry

  @inproceedings{jain2022missingness,
      title={Missingness Bias in Model Debugging},
      author={Saachi Jain and Hadi Salman and Eric Wong and Pengchuan Zhang and Vibhav Vineet and Sai Vemprala and Aleksander Madry},
      booktitle={International Conference on Learning Representations},
      year={2022},
      url={https://openreview.net/forum?id=Te5ytkqsnl}
    }

All scripts to run the experiments in the paper are in patch_ablation/.

Patch Ablation Experiments (Section 3)

To generate the patch ablation experiments, run the script run_patch_ablation.py. For example, to run the basic patch ablation experiments for a ResNet50:

python run_patch_ablation.py --arch resnet50 --out-dir OUT_DIR --methods Saliency_reverse --methods Saliency --ablation-patch-size 16 --saliency-map-pkl model_checkpoints/resnet_50_sal_map.pkl --filler-values 0 0 0 --skip-factor 2

and for a ViT-S:

python run_patch_ablation.py --arch deit_small_resnet_aug --out-dir OUT_DIR --methods Saliency_reverse --methods Saliency --ablation-patch-size 16 --saliency-map-pkl  model_checkpoints/resnet_50_sal_map.pkl --filler-values 0 0 0 --skip-factor 2 --use-missingness

Generate LIME for a given image (Section 4)

Run run_lime.py to generate LIME explanations for a given image (using either blacking out or dropping tokens for ViTs). Example:

python run_lime.py --out-dir  OUTDIR --ablation-patch-size 16 --filler-values 0 0 0 --superpixel-type patches --arch resnet50 --num-perturbations 1000 --batch-size 64 

python run_lime.py --out-dir  OUTDIR --ablation-patch-size 16 --filler-values 0 0 0 --superpixel-type patches --arch deit_small_resnet_aug --num-perturbations 1000 --batch-size 64 --use-missingness

Perform top-K ablation test (Section 4)

Run run_lime_ablation.py to perform the top K ablation test on LIME explanations. Example:

python run_lime_ablation.py --model-name resnet50 --out-pkl OUT_PKL.pkl --lime-pkl LIME_PKL --ablation-patch-size 16 --num-features 50 --feature-skip 2 --superpixel-type patches

python run_lime_ablation.py --model-name deit_small_resnet_aug_missingness --out-pkl OUT_PKL.pkl --lime-pkl LIME_PKL --ablation-patch-size 16 --num-features 50 --feature-skip 2 --superpixel-type patches

Download our Models and Data

We further have the following files on dropbox since they were too large:

Have further questions?

Feel free to raise issues or contact us!