Controlled benchmarks show INR architectures differ in whether weight reuse is source-specific or generic, with no architecture dominating all PDE and analytic cases.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Architecture Shapes Transfer Specificity in Implicit Neural Representations
Controlled benchmarks show INR architectures differ in whether weight reuse is source-specific or generic, with no architecture dominating all PDE and analytic cases.