Search Results for author:
Found 2 papers, 0 papers with code
Train Faster, Perform Better: Modular Adaptive Training in Over-Parameterized Models
To describe such modular-level learning capabilities, we introduce a novel concept dubbed modular neural tangent kernel (mNTK), and we demonstrate that the quality of a module's learning is tightly associated with its mNTK's principal eigenvalue $\lambda_{\max}$.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
