Auto-conditioned Frank-Wolfe methods use local Lipschitz estimators from first-order information to achieve convergence guarantees in convex and nonconvex settings without prior global smoothness knowledge.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
method 1
citation-polarity summary
verdicts
UNVERDICTED 3roles
method 1polarities
use method 1representative citing papers
CrackMorph-XAI-Net extracts crack skeletons with Dice 0.991 and topology preservation in 98.5% of cases, detects junctions with F1 0.887, and computes morphology descriptors with correlations above 0.95 on an extended CRACK500 benchmark.
Proposes low-rank orthogonalization and derives low-rank Muon and MSGD variants that outperform standard Muon on GPT-2 and LLaMA pretraining while providing iteration complexity bounds.
citing papers explorer
-
Auto-Conditioned Frank-Wolfe Algorithms
Auto-conditioned Frank-Wolfe methods use local Lipschitz estimators from first-order information to achieve convergence guarantees in convex and nonconvex settings without prior global smoothness knowledge.
-
CrackMorph-XAI-Net: A Topology-Preserving and Explainable Framework for Automated Crack Morphology
CrackMorph-XAI-Net extracts crack skeletons with Dice 0.991 and topology preservation in 98.5% of cases, detects junctions with F1 0.887, and computes morphology descriptors with correlations above 0.95 on an extended CRACK500 benchmark.
-
Low-rank Orthogonalization for Large-scale Matrix Optimization with Applications to Foundation Model Training
Proposes low-rank orthogonalization and derives low-rank Muon and MSGD variants that outperform standard Muon on GPT-2 and LLaMA pretraining while providing iteration complexity bounds.