Unverified50% confidenceFactExact time
GCN's degree-normalization coefficients depend on node degree, and deleting an edge changes the degree and thus the coefficients of the surviving edges, so GCN does not satisfy the equivalence condition of "edge coefficients independent of the edge set."
1
Sources
50%
Confidence
Long-term
Relevance
11/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedFor GNN models with additive message passing and edge coefficients independent of the edge set, structural deletion and message suppression are exactly equivalent interventions.71% similarUnverifiedOn the feeder data, parameter sharing recovers about two-thirds of the error gap between node-specific models and the GNN (relative to the regularized baseline).70% similarUnverifiedWhen comparing different GNN explanation methods, if each uses a different edge-deletion semantics, the comparison may not be conducted on the same basis.70% similarUnverifiedOn the target logit, 10 of the 11 reportable GCN and GATv2 units show relative fidelity that depends on the evaluation intervention; in 9 of these units, each ranking achieves a larger deletion area under the intervention it was itself constructed for.67% similarUnverifiedGNNs address the problems in graph data of a non-fixed number of neighbors and meaningless ordering through the message passing mechanism: in each layer, a node first aggregates the feature information of its neighbors, then combines it with its own representation and updates it through a nonlinear transformation.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1134408API
curl https://kongchang.com/api/v1/knowledge/claims/1134408MCP
get_claim(id=1134408)