Pith. sign in

REVIEW 1 cited by

Revisiting Sanity Checks for Saliency Maps

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2110.14297 v1 pith:6HFYE6LQ submitted 2021-10-27 cs.LG cs.CVcs.HC

Revisiting Sanity Checks for Saliency Maps

classification cs.LG cs.CVcs.HC
keywords saliencychallengemapsmethodssanitytaskstheyargue
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Saliency methods are a popular approach for model debugging and explainability. However, in the absence of ground-truth data for what the correct maps should be, evaluating and comparing different approaches remains a long-standing challenge. The sanity checks methodology of Adebayo et al [Neurips 2018] has sought to address this challenge. They argue that some popular saliency methods should not be used for explainability purposes since the maps they produce are not sensitive to the underlying model that is to be explained. Through a causal re-framing of their objective, we argue that their empirical evaluation does not fully establish these conclusions, due to a form of confounding introduced by the tasks they evaluate on. Through various experiments on simple custom tasks we demonstrate that some of their conclusions may indeed be artifacts of the tasks more than a criticism of the saliency methods themselves. More broadly, our work challenges the utility of the sanity check methodology, and further highlights that saliency map evaluation beyond ad-hoc visual examination remains a fundamental challenge.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. RoentMod: A Synthetic Chest X-Ray Modification Model to Identify and Correct Image Interpretation Model Shortcuts

    eess.IV 2025-09 conditional novelty 6.0

    RoentMod edits chest X-rays to add specific diseases, exposing and partly correcting shortcut learning in AI interpretation models.