5 papers
Mixture-of-Depths Attention
Lianghui Zhu, Yuxin Fang, Bencheng Liao +10
Scaling depth is a key driver for large language models (LLMs). Yet, as LLMs become deeper, they often suffer from signal degradation: informative features formed in shallow layers…
HybridNorm: Towards Stable and Efficient Transformer Training via Hybrid Normalization
Zhijian Zhuo, Yutao Zeng, Ya Wang +5
Transformers have become the de facto architecture for a wide range of machine learning tasks, particularly in large language models (LLMs). Despite their remarkable performance, m…
Over-Tokenized Transformer: Vocabulary is Generally Worth Scaling
Hongzhi Huang, Defa Zhu, Banggu Wu +4
Tokenization is a fundamental component of large language models (LLMs), yet its influence on model scaling and performance is not fully explored. In this paper, we introduce Over-…
Efficient Pretraining Length Scaling
Bohong Wu, Shen Yan, Sijun Zhang +4
Recent advances in large language models have demonstrated the effectiveness of length scaling during post-training, yet its potential in pre-training remains underexplored. We pre…
Scale-Distribution Decoupling: Enabling Stable and Effective Training of Large Language Models
Ya Wang, Zhijian Zhuo, Yutao Zeng +3
Training stability is a persistent challenge in the pre-training of large language models (LLMs), particularly for architectures such as Post-Norm Transformers, which are prone to…