A denoising auxiliary loss on perturbed entity embeddings improves KGE robustness on FB15k-237, but the proposed certified robustness metrics rest on a misapplication of randomized smoothing.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Robust Knowledge Graph Embedding via Denoising
A denoising auxiliary loss on perturbed entity embeddings improves KGE robustness on FB15k-237, but the proposed certified robustness metrics rest on a misapplication of randomized smoothing.