Skip to content

Conversation

@aobo-y
Copy link
Contributor

@aobo-y aobo-y commented Oct 30, 2023

Summary: as title

Reviewed By: vivekmig

Differential Revision: D50715178

@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D50715178

Summary:

Support multi-task attribution in `ShapleyValues` and `ShapleyValueSampling`.

Assuming the return of `forward_fun` is in (*output_shape), the attribution result will be in (*output_shape, *input_shape[1:]). Existing use cases becomes just special cases where output_shape is (1,) or (batch_size,)

Reviewed By: vivekmig

Differential Revision: D48696578
Summary:

As ShapleyValues and SHapleyValueSampling have supported multi-task output, they can be used with LLMAttribution

Reviewed By: vivekmig

Differential Revision: D48697749
Summary:

as title

Reviewed By: vivekmig

Differential Revision: D50715178
aobo-y added a commit to aobo-y/captum that referenced this pull request Oct 31, 2023
Summary:

as title

Reviewed By: vivekmig

Differential Revision: D50715178
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D50715178

@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D50715178

@facebook-github-bot
Copy link
Contributor

This pull request has been merged in e6ee896.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants