arXiv Machine Learning
Jul 24
Overcoming the Communication-Performance Tradeoff in LLM Pretraining
arXiv:2508. 15706v3 Announce Type: replace Abstract: Communication-efficient distributed training algorithms (e.
SCAPE: Accurate and Efficient LLM Training with Extreme Sparse Communication
arXiv:2607. 01678v1 Announce Type: new Abstract: Communication increasingly dominates the cost of Large Language Model (LLM) pre-training, especially under data-parallel and sharded training schemes, where gradient synchronization and parameter reconstruction overhead increase with model size and system scale.
Related stories
arXiv AI
Jun 12
MiniMax Sparse Attention
arXiv:2606. 13392v1 Announce Type: new Abstract: Ultra-long-context capability is becoming indispensable for frontier LLMs: agentic workflows, repository-scale code reasoning, and persistent memory all require the model to jointly attend over hundreds of thousands to millions of tokens, yet the quadratic cost of softmax attention makes this untenable at deployment scale.
arXiv Machine Learning
Jun 4
SparDA: Sparse Decoupled Attention for Efficient Long-Context LLM Inference
arXiv:2606. 04511v1 Announce Type: cross Abstract: Sparse attention reduces compute and memory bandwidth for long-context LLM inference.
arXiv AI
Jun 19
LoRDO: Distributed Low-Rank Optimization with Infrequent Communication
arXiv:2602. 04396v2 Announce Type: replace-cross Abstract: Distributed training of foundation models via $\texttt{DDP}$ is limited by interconnect bandwidth.
arXiv Machine Learning
Jun 10
AdaGC: Enhancing LLM Pretraining Stability via Adaptive Gradient Clipping
arXiv:2502. 11034v3 Announce Type: replace Abstract: Loss spikes remain a persistent obstacle in large-scale language model pretraining.
arXiv AI
Jun 16
Communication-Efficient Verifiable Attention for LLM Inference
arXiv:2606. 16352v1 Announce Type: cross Abstract: Computation integrity of remote large language model (LLM) serving can be questionable.
arXiv Machine Learning
Jun 9
Efficient Scaling of LLM Training with Flexible Context Parallelism
arXiv:2602. 21788v2 Announce Type: replace-cross Abstract: Scaling long-context capabilities is crucial for Large Language Models (LLMs).
arXiv AI
Jun 2
When Data Is Scarce: Scaling Sparse Language Models with Repeated Training
arXiv:2606. 01155v1 Announce Type: cross Abstract: Scaling laws for dense LLMs under infinite data are well explored, but how sparsity interacts with limited data is not.
arXiv AI
Jul 24
SOAP, Muon, and Beyond: Pushing LLM Pretraining Scales
arXiv:2607. 20548v1 Announce Type: cross Abstract: Higher-order optimizers such as Muon and SOAP offer faster convergence than AdamW, but their computational cost and numerical stability challenges have limited adoption at scale.
arXiv Machine Learning
Jun 18
Mixed-Precision Communication-Avoiding SGD for Generalized Linear Models on GPUs
arXiv:2606. 18463v1 Announce Type: cross Abstract: Distributed stochastic gradient descent (SGD) is limited by communication rather than computation, since each iteration requires an AllReduce across processes.
arXiv AI
Jun 10
Unifying Local Communications and Local Updates for LLM Pretraining
arXiv:2606. 11081v1 Announce Type: cross Abstract: Communication-efficient pre-training of LLMs is increasingly important as training draws on compute distributed across clusters, data centers, and lower-bandwidth links.
arXiv Machine Learning
Jun 25
EPTS: Elastic Post-Training Sparsity for Efficient Large Language Model Compression
arXiv:2606. 25285v1 Announce Type: new Abstract: Post-Training Sparsity (PTS) has emerged as a crucial paradigm for compressing Large Language Models to facilitate efficient deployment on resource-constrained devices.