efficient training 1large language models 1residual stream expansion 1sparse computation 1transformer architectures 1
From the 1 of 3 linked papers with an AI index.
3 papers
cs.LG2026
xHC: Expanded Hyper-Connections
Xiangdong Zhang, Xiaohan Qin, Sunan Zou +10
The paper introduces xHC, a method that expands the residual stream of Transformers to many parallel streams using temporal feature augmentation and a sparse update scheme, enablin…
cs.LG2026
BigMac: Breaking the Pareto Frontier of Compute and Memory in Multimodal LLM Training
Zili Zhang, Chengxu Yang, Shenglong Zhang +8
Training multimodal large language models (MLLMs) is challenged by both model and data heterogeneity. Existing systems redesign the training pipeline to address these challenges, b…
cs.DC2026
UltraEP: Unleash MoE Training and Inference on Rack-Scale Nodes with Near-Optimal Load Balancing
Xinming Wei, Chao Jin, Tuo Dai +10
Large-scale expert parallelism (EP) is becoming pivotal for training and serving frontier MoE models, but it also amplifies device-level expert load imbalance into compute straggle…