HybridNorm: Towards Stable and Efficient Transformer Training via Hybrid Normalization Paper • 2503.04598 • Published Mar 6 • 21
Over-Tokenized Transformer: Vocabulary is Generally Worth Scaling Paper • 2501.16975 • Published Jan 28 • 31
Polynomial Composition Activations: Unleashing the Dynamics of Large Language Models Paper • 2411.03884 • Published Nov 6, 2024 • 28
