5 papers
Virtual Width Networks
Seed, Baisheng Li, Banggu Wu +115
We introduce Virtual Width Networks (VWN), a framework that delivers the benefits of wider representations without incurring the quadratic cost of increasing the hidden size. VWN d…
Enhancing Linear Attention with Residual Learning
Xunhao Lai, Jialiang Kang, Jianqiao Lu +2
Linear attention offers a linear-time alternative to self-attention but often struggles to capture long-range patterns. We revisit linear attention through a prediction-correction…
Model Merging in Pre-training of Large Language Models
Yunshui Li, Yiyuan Ma, Shen Yan +23
Model merging has emerged as a promising technique for enhancing large language models, though its application in large-scale pre-training remains relatively unexplored. In this pa…
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…
FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference
Xunhao Lai, Jianqiao Lu, Yao Luo +2
Large language models (LLMs) encounter computational challenges during long-sequence inference, especially in the attention pre-filling phase, where the complexity grows quadratica…