DAM: Dynamic Attention Mask for Long-Context Large Language Model Inference Acceleration

• URL: http://arxiv.org/abs/2506.11104v1
• Date: Fri, 06 Jun 2025 20:24:36 GMT
• Title: DAM: Dynamic Attention Mask for Long-Context Large Language Model Inference Acceleration
• Authors: Hanzhi Zhang, Heng Fan, Kewei Sha, Yan Huang, Yunhe Feng,
• Abstract summary: We introduce a dynamic sparse attention mechanism that assigns adaptive masks at the attention-map level.<n>By learning context-aware attention structures, it achieves high alignment with full-attention models, ensuring minimal performance degradation.<n>This approach provides a scalable alternative to full attention, enabling the practical deployment of large-scale Large Language Models.
• Score: 12.172968576254469
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Long-context understanding is crucial for many NLP applications, yet transformers struggle with efficiency due to the quadratic complexity of self-attention. Sparse attention methods alleviate this cost but often impose static, predefined masks, failing to capture heterogeneous attention patterns. This results in suboptimal token interactions, limiting adaptability and retrieval accuracy in long-sequence tasks. This work introduces a dynamic sparse attention mechanism that assigns adaptive masks at the attention-map level, preserving heterogeneous patterns across layers and heads. Unlike existing approaches, our method eliminates the need for fine-tuning and predefined mask structures while maintaining computational efficiency. By learning context-aware attention structures, it achieves high alignment with full-attention models, ensuring minimal performance degradation while reducing memory and compute overhead. This approach provides a scalable alternative to full attention, enabling the practical deployment of large-scale Large Language Models (LLMs) without sacrificing retrieval performance. DAM is available at: https://github.com/HanzhiZhang-Ulrica/DAM.

Related papers

• Training-free Context-adaptive Attention for Efficient Long Context Modeling [57.703159205740185]
  Training-free Context-adaptive Attention (TCA-Attention) is a training-free sparse attention mechanism that selectively attends to only the informative tokens for efficient long-context inference.<n>TCA-Attention achieves a 2.8$times$ speedup and reduces KV cache by 61% at 128K context length while maintaining performance comparable to full attention.
  arXiv  Detail & Related papers  (2025-12-10T01:54:57Z)
• PSA: Pyramid Sparse Attention for Efficient Video Understanding and Generation [34.8993443618652]
  We present Pyramid Sparse Attention (PSA), a versatile module applicable to both video understanding and generation tasks.<n>Instead of binary masking, PSA introduces multi-level pooled KV representations, enabling finer mask granularity.<n>This design, analogous to fixed-point quantization and classical feature pyramid networks in computer vision, effectively mitigates information loss while preserving computational efficiency under a low compute budget.
  arXiv  Detail & Related papers  (2025-12-03T18:02:11Z)
• InfLLM-V2: Dense-Sparse Switchable Attention for Seamless Short-to-Long Adaptation [56.694702609077495]
  Long-sequence processing is a critical capability for modern large language models.<n>InfLLM-V2 is a trainable sparse attention framework that seamlessly adapts models from short to long sequences.<n>In experiments, InfLLM-V2 is 4$times$ faster than dense attention while retaining 98.1% and 99.7% of the performance.
  arXiv  Detail & Related papers  (2025-09-29T12:08:33Z)
• Trainable Dynamic Mask Sparse Attention [11.506985057671015]
  We introduce a trainable dynamic mask sparse attention mechanism, a method that merges the advantages of both position-aware and content-aware approaches.<n>We demonstrate that the introduced dynamic mask and sparse weights do not obstruct gradients, supporting end-to-end training.
  arXiv  Detail & Related papers  (2025-08-04T07:05:15Z)
• Modality Agnostic Efficient Long Range Encoder [14.705955027331674]
  We address the challenge of long-context processing on a single device using generic implementations.<n>To overcome these limitations, we propose MAELRE, a unified and efficient transformer architecture.<n>We demonstrate that MAELRE achieves superior accuracy while reducing computational cost compared to existing long-context models.
  arXiv  Detail & Related papers  (2025-07-25T16:19:47Z)
• Attention, Please! Revisiting Attentive Probing for Masked Image Modeling [20.39513629593113]
  We introduce efficient probing (EP), a cross-attention mechanism that eliminates redundant projections, reduces the number of trainable parameters, and achieves up to a 10$times$ speed-up over conventional multi-head attention.<n>EP generalizes well beyond MIM to diverse pre-training paradigms, produces interpretable attention maps, and achieves strong gains in low-shot and layer-wise settings.
  arXiv  Detail & Related papers  (2025-06-11T21:10:26Z)
• ALTo: Adaptive-Length Tokenizer for Autoregressive Mask Generation [6.618610740026886]
  We propose ALTo, an adaptive length tokenizer for autoregressive mask generation.<n>A novel token length predictor is designed, along with a length regularization term and a differentiable token chunking strategy.<n>Experiments demonstrate that ALToLLM achieves state-of-the-art performance with adaptive token cost.
  arXiv  Detail & Related papers  (2025-05-22T10:26:51Z)
• Sliding Window Attention Training for Efficient Large Language Models [55.56483740523027]
  We introduce SWAT, which enables efficient long-context handling via Sliding Window Attention Training.<n>This paper first attributes the inefficiency of Transformers to the attention sink phenomenon.<n>We replace softmax with the sigmoid function and utilize a balanced ALiBi and Rotary Position Embedding for efficient information compression and retention.
  arXiv  Detail & Related papers  (2025-02-26T05:31:44Z)
• Taipan: Efficient and Expressive State Space Language Models with Selective Attention [100.16383527459429]
  Long-context language modeling is a significant challenge in Natural Language Processing (NLP) Recent State Space Models (SSMs) such as Mamba offer alternatives with constant memory usage, but they underperform in tasks requiring extensive in-context retrieval. We introduce Taipan, a novel hybrid architecture that combines Mamba-2 with Selective Attention Layers (SALs) Our experiments demonstrate Taipan's superior performance across various scales and tasks, offering a promising solution for efficient long-context language modeling.
  arXiv  Detail & Related papers  (2024-10-24T09:25:37Z)
• ColorMAE: Exploring data-independent masking strategies in Masked AutoEncoders [53.3185750528969]
  Masked AutoEncoders (MAE) have emerged as a robust self-supervised framework. We introduce a data-independent method, termed ColorMAE, which generates different binary mask patterns by filtering random noise. We demonstrate our strategy's superiority in downstream tasks compared to random masking.
  arXiv  Detail & Related papers  (2024-07-17T22:04:00Z)
• Sparser is Faster and Less is More: Efficient Sparse Attention for Long-Range Transformers [58.5711048151424]
  We introduce SPARSEK Attention, a novel sparse attention mechanism designed to overcome computational and memory obstacles. Our approach integrates a scoring network and a differentiable top-k mask operator, SPARSEK, to select a constant number of KV pairs for each query. Experimental results reveal that SPARSEK Attention outperforms previous sparse attention methods.
  arXiv  Detail & Related papers  (2024-06-24T15:55:59Z)
• CItruS: Chunked Instruction-aware State Eviction for Long Sequence Modeling [52.404072802235234]
  We introduce Chunked Instruction-aware State Eviction (CItruS), a novel modeling technique that integrates the attention preferences useful for a downstream task into the eviction process of hidden states. Our training-free method exhibits superior performance on long sequence comprehension and retrieval tasks over several strong baselines under the same memory budget.
  arXiv  Detail & Related papers  (2024-06-17T18:34:58Z)
• A Training-free Sub-quadratic Cost Transformer Model Serving Framework With Hierarchically Pruned Attention [43.211427581302715]
  We propose Hierarchically Pruned Attention (HiP) to increase context length in large language models.<n>HiP reduces the time complexity of the attention mechanism to $O(T log T)$ and the space complexity to $O(T)$, where $T$ is the sequence length.<n>We show that HiP significantly reduces both prefill and decoding latencies, as well as memory usage, while maintaining high-quality generation with minimal degradation.
  arXiv  Detail & Related papers  (2024-06-14T08:32:45Z)
• LongVQ: Long Sequence Modeling with Vector Quantization on Structured Memory [63.41820940103348]
  Self-attention mechanism's computational cost limits its practicality for long sequences. We propose a new method called LongVQ to compress the global abstraction as a length-fixed codebook. LongVQ effectively maintains dynamic global and local patterns, which helps to complement the lack of long-range dependency issues.
  arXiv  Detail & Related papers  (2024-04-17T08:26:34Z)
• Efficient Long-Range Transformers: You Need to Attend More, but Not Necessarily at Every Layer [36.75562615596186]
  We propose MASFormer, an easy-to-implement transformer variant with Mixed Attention Spans. MASFormer is equipped with full attention to capture long-range dependencies, but only at a small number of layers. Experiments show that a decoder-only MASFormer model of 1.3B parameters can achieve competitive performance to vanilla transformers with full attention.
  arXiv  Detail & Related papers  (2023-10-19T03:32:05Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.