3 papers across 3 sessions
This paper explores the generalization mechanism of KANs and designs more effective KANs with lower model complexity and better generalization.