Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Wrong results for log standard error #32

Open
statkwon opened this issue Aug 17, 2024 · 0 comments
Open

Wrong results for log standard error #32

statkwon opened this issue Aug 17, 2024 · 0 comments

Comments

@statkwon
Copy link

Thanks for providing wonderful benchmark for XAI methods.

But there's a slight issue in the provided codes. When I evaluate metrics using evaluate_metrics.py, I cannot get the accurate results for log standard error.

For example, the scale of log standard error for RIS below seems awkward.

Model: lr, Data: adult, Explainer: control (1/16)1s total, 1s on model, 1s on dataset)
PRA: 0.501 ± 0.003

RC: 0.003 ± 0.009

FA: 0.188 ± 0.006

RA: 0.073 ± 0.005

SA: 0.093 ± 0.004

SRA: 0.034 ± 0.003

PGU: 0.130 ± 0.003

PGI: 0.043 ± 0.002

RIS: 420764.219 ± 1391.567
log(RIS): 12.950 ± 7.238

I made PR #31 to fix this issue. Please check this.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant