Pith. sign in

REVIEW

The effectiveness of feature attribution methods and its correlation with automatic evaluation scores

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 2105.14944 v5 pith:EKQLNNCS submitted 2021-05-31 cs.CV cs.AIcs.HC

The effectiveness of feature attribution methods and its correlation with automatic evaluation scores

classification cs.CV cs.AIcs.HC
keywords attributionhumansmethodsevaluationfeatureadversarialapplicationsautomatic
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Explaining the decisions of an Artificial Intelligence (AI) model is increasingly critical in many real-world, high-stake applications. Hundreds of papers have either proposed new feature attribution methods, discussed or harnessed these tools in their work. However, despite humans being the target end-users, most attribution methods were only evaluated on proxy automatic-evaluation metrics (Zhang et al. 2018; Zhou et al. 2016; Petsiuk et al. 2018). In this paper, we conduct the first user study to measure attribution map effectiveness in assisting humans in ImageNet classification and Stanford Dogs fine-grained classification, and when an image is natural or adversarial (i.e., contains adversarial perturbations). Overall, feature attribution is surprisingly not more effective than showing humans nearest training-set examples. On a harder task of fine-grained dog categorization, presenting attribution maps to humans does not help, but instead hurts the performance of human-AI teams compared to AI alone. Importantly, we found automatic attribution-map evaluation measures to correlate poorly with the actual human-AI team performance. Our findings encourage the community to rigorously test their methods on the downstream human-in-the-loop applications and to rethink the existing evaluation metrics.

discussion (0)

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