Metraipolis › archive › dc_7cefe1b35b6d

What Does It Mean to Remove an Edge? Intervention-Aware Fidelity for Graph Neural Network Explanations — arXiv:2611.04751

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.04751
License (as archived)
CC0-1.0
Archived
2026-11-04T18:54:12.152Z
SHA-256 of the archived text
bd1d3e1f0d139ca79f89cdb34b2c89fd2f6eec2ff3e00d5c1f86e0e55bb1fe0d
Stable URL of this page
https://metraipolis.com/documento/dc_7cefe1b35b6d

How to cite

What Does It Mean to Remove an Edge? Intervention-Aware Fidelity for Graph Neural Network Explanations — arXiv:2611.04751 — arXiv (en), https://arxiv.org/abs/2611.04751 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7cefe1b35b6d (Metraipolis archive dc_7cefe1b35b6d, sha256 bd1d3e1f0d139ca79f89cdb34b2c89fd2f6eec2ff3e00d5c1f86e0e55bb1fe0d)
BibTeX
@misc{dc7cefe1b35b6d,
  title = {What Does It Mean to Remove an Edge? Intervention-Aware Fidelity for Graph Neural Network Explanations — arXiv:2611.04751},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04751},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7cefe1b35b6d; sha256 bd1d3e1f0d139ca79f89cdb34b2c89fd2f6eec2ff3e00d5c1f86e0e55bb1fe0d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04751. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Graph neural network explanations are often evaluated via deletion fidelity, i.e., by deleting their top-ranked edges and measuring the resulting change in model output. Yet, in a message-passing GNN, "deleting an edge" does not specify a unique intervention. Structural deletion removes the edge from the graph and recomputes quantities such as degree normalization and attention, whereas message suppression leaves this structural context intact while blocking the message carried by the edge. The meaning of deletion fidelity therefore depends on what the intervention targets in the model computation. We formalize this distinction and derive sufficient conditions for exact equivalence. The two interventions coincide for additive message passing with edge-set-independent coefficients, but can diverge when aggregation depends on the presence of zero messages or when removing an edge changes the coefficients applied to surviving messages. We then construct intervention-specific edge attributions and evaluate each resulting ranking under both interventions, testing whether their relative fidelity changes with the evaluation intervention. Across nine datasets and a panel of GIN, GCN, and GATv2 models at depths 2--5, all nine GIN controls satisfy the equivalence predicted by the theory. On the target logit, the relative fidelity of the two rankings depends on the evaluation intervention in 10 of 11 reportable GCN and GATv2 cells, and in nine of them each ranking has the larger deletion area under the intervention it was built with. Deletion fidelity is thus not a property of an edge ranking alone. Evaluation protocols should explicitly report the intervention, output functional, target, and perturbation unit, and test alternative intervention semantics when the scientific question does not uniquely determine one.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.04751 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-04T18:35:37.456Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).