↓Skip to main content

🏢 CUHK

DAPE: Data-Adaptive Positional Encoding for Length Extrapolation
·3365 words·16 mins· loading · loading
Natural Language Processing Large Language Models 🏢 CUHK
DAPE: A novel data-adaptive positional encoding method dynamically adjusts positional information based on input context, improving transformer performance and length generalization.