# LASP-2: Rethinking Sequence Parallelism for Linear Attention and Its Hybrid

arXiv: https://arxiv.org/abs/2502.07563 · HTML: https://mlsanity.com/paper/2502.07563/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Feb 2025.

## Built on by (most cited first, 9 of 9)

- [Kimi K3: Open Frontier Intelligence](https://mlsanity.com/paper/2607.24653/) (2026)
- [A Survey of Efficient Reasoning for Large Reasoning Models: Language, Multimodality, and Beyond](https://mlsanity.com/paper/2503.21614/) (2025)
- [MoM: Linear Sequence Modeling with Mixture-of-Memories](https://mlsanity.com/paper/2502.13685/) (2025)
- [Liger: Linearizing Large Language Models to Gated Recurrent Structures](https://mlsanity.com/paper/2503.01496/) (2025)
- [Native Hybrid Attention for Efficient Sequence Modeling](https://mlsanity.com/paper/2510.07019/) (2025)
- [Linear-MoE: Linear Sequence Modeling Meets Mixture-of-Experts](https://mlsanity.com/paper/2503.05447/) (2025)
- [SpikingBrain: Spiking Brain-inspired Large Models](https://mlsanity.com/paper/2509.05276/) (2025)
- [ZeCO: Zero Communication Overhead Sequence Parallelism for Linear Attention](https://mlsanity.com/paper/2507.01004/) (2025)
- [Fold-CP: A Context Parallelism Framework for Biomolecular Modeling](https://mlsanity.com/paper/2603.14806/) (2026)

## Abstract (arXiv)

Linear sequence modeling approaches, such as linear attention, provide advantages like linear-time training and constant-memory inference over sequence lengths. However, existing sequence parallelism (SP) methods are either not optimized for the right-product-first feature of linear attention or use a ring-style communication strategy, which results in lower computation parallelism, limits their scalability for longer sequences in distributed systems. In this paper, we introduce LASP-2, a new SP method to enhance both communication and computation parallelism when training linear attention transformer models with very-long input sequences. Compared to previous work LASP, LASP-2 rethinks the minimal communication requirement for SP on linear attention layers, reorganizes the whole communication-computation workflow of LASP. In this way, only one single AllGather collective communication is needed on intermediate memory states, whose sizes are independent of the sequence length, leading to significant improvements of both communication and computation parallelism, as well as their overlap. Additionally, we extend LASP-2 to LASP-2H by applying similar communication redesign to standard attention modules, offering an efficient SP solution for hybrid models that blend linear and standard attention layers. Our evaluation on a Linear-Llama3 model, a variant of Llama3 with linear attention replacing standard attention, demonstrates the effectiveness of LASP-2 and LASP-2H. Specifically, LASP-2 achieves training speed improvements of 15.2% over LASP and 36.6% over Ring Attention, with a sequence length of 2048K across 64 GPUs. The Code is released as a part of: https://github.com/OpenSparseLLMs/Linear-MoE.