Document PyG graph model attribution workflow - #1862
Open
craymichael wants to merge 1 commit into
Open
Conversation
Fixes meta-pytorch#246. Issue: meta-pytorch#246 Issue summary: Users asked how to use Captum with PyTorch Geometric / graph neural network models and whether Captum needs graph-native support. Summary: - Adds an FAQ entry for PyTorch Geometric and graph neural network attribution workflows. - Points users to PyG's CaptumExplainer, to_captum_model, and to_captum_input utilities for graph-native node / edge masks. - Shows the lower-level Captum pattern for tensor node-feature attribution with graph structure passed as additional_forward_args. Test Plan: - git diff --check - Documentation-only change; no runtime tests run.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #246.
Summary
CaptumExplainer,to_captum_model, andto_captum_input.edge_index/batchasadditional_forward_args.edge_indexis an integer topology tensor and should not be attributed directly; edge or topology importance should use an edge mask, edge weight, or edge feature tensor representation.Test Plan
git diff --check