Scaling vision models by depth and parameter count does not consistently improve localisation-based explanation quality across architectures, datasets, and post-hoc methods; smaller models often perform comparably or better.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
Dense ReLU networks under natural weight and dimension constraints fail to approximate certain Lipschitz functions, unlike unrestricted networks.
Energy-aware metaheuristics use an EI/J score to dynamically pick operators that maximize fitness gain per unit energy, reaching comparable fitness with substantially less energy than standard versions on knapsack, NK-landscapes, and error-correcting code problems.
citing papers explorer
-
Scaling Vision Models Does Not Consistently Improve Localisation-Based Explanation Quality
Scaling vision models by depth and parameter count does not consistently improve localisation-based explanation quality across architectures, datasets, and post-hoc methods; smaller models often perform comparably or better.
-
Neural Networks With Dense Weights Are Not Universal Approximators
Dense ReLU networks under natural weight and dimension constraints fail to approximate certain Lipschitz functions, unlike unrestricted networks.
-
Energy-Aware Metaheuristics
Energy-aware metaheuristics use an EI/J score to dynamically pick operators that maximize fitness gain per unit energy, reaching comparable fitness with substantially less energy than standard versions on knapsack, NK-landscapes, and error-correcting code problems.