MDIR detects LLM weight homology from embedding matrices alone using polar decomposition and permutation matching, achieving perfect AUC and accuracy on LeaFBench and reconstructing layer-level transformations.
Canonical fields: Self-supervised learning of pose-canonicalized neural fields
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
FedMeNF: Privacy-Preserving Federated Meta-Learning for Neural Fields
MDIR detects LLM weight homology from embedding matrices alone using polar decomposition and permutation matching, achieving perfect AUC and accuracy on LeaFBench and reconstructing layer-level transformations.