[ICML 2024] Outlier-Efficient Hopfield Layers for Large Transformer-Based Models
transformer outliers attention attention-mechanism outlier-removal outlier hopfield-neural-network ptq outlier-treatment modern-hopfield-networks modern-hopfield-model icml-2024 softmax-1 quantized-friendly no-op-outlier
-
Updated
Oct 17, 2024 - Python