-
Notifications
You must be signed in to change notification settings - Fork 1.2k
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
Fixed non-scalar (text, vector, set) output feature explanations #3269
Conversation
# If the target feature is a non-scalar type (vector, set, etc.), sum it to get a scalar value. | ||
# https://github.com/pytorch/captum/issues/377 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
QQ: Does this change the interpretation/explanation for these kinds of features?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It would be different from looking at each element of the output individually, yes, but it's difficult to reason about what that alternative means for things like vectors.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Seems reasonable to me! Thanks for finding and linking to a relevant discussion on the captum project.
Unit Test Results 5 files - 1 5 suites - 1 17m 33s ⏱️ - 1h 32m 10s Results for commit 29da637. ± Comparison against base commit 6fc4713. This pull request removes 146 tests.
♻️ This comment has been updated with latest results. |
No description provided.