Headless Language Models: Learning Without Predicting with Contrastive Weight TyingNathan Godey,Éric Villemonte de la Clergerie,Benoît SagotICLR 2024(2024)引用 2|浏览66关键词representation learning,NLP,language modeling,pretraining,contrastiveAI 理解论文溯源树样例生成溯源树,研究论文发展脉络Chat Paper正在生成论文摘要