Switch Attention: Towards Dynamic and Fine-grained Hybrid Transformers

• URL: http://arxiv.org/abs/2603.26380v1
• Date: Fri, 27 Mar 2026 13:04:29 GMT
• Title: Switch Attention: Towards Dynamic and Fine-grained Hybrid Transformers
• Abstract summary: We propose a novel hybrid transformer that enables dynamic and fine-grained routing between full attention and sliding window attention.<n>SwiAttn routes dynamically the computation to either a full-attention branch for global information aggregation or a sliding-window branch for efficient local pattern matching.<n>Experiments are conducted on twenty-three benchmark datasets across both regular (4K) and long (32K) context lengths.
• Score: 34.75985836977073
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: The attention mechanism has been the core component in modern transformer architectures. However, the computation of standard full attention scales quadratically with the sequence length, serving as a major bottleneck in long-context language modeling. Sliding window attention restricts the context length for better efficiency at the cost of narrower receptive fields. While existing efforts attempt to take the benefits from both sides by building hybrid models, they often resort to static, heuristically designed alternating patterns that limit efficient allocation of computation in various scenarios. In this paper, we propose Switch Attention (SwiAttn), a novel hybrid transformer that enables dynamic and fine-grained routing between full attention and sliding window attention. For each token at each transformer layer, SwiAttn dynamically routes the computation to either a full-attention branch for global information aggregation or a sliding-window branch for efficient local pattern matching. An adaptive regularization objective is designed to encourage the model towards efficiency. Moreover, we adopt continual pretraining to optimize the model, transferring the full attention architecture to the hybrid one. Extensive experiments are conducted on twenty-three benchmark datasets across both regular (4K) and long (32K) context lengths, demonstrating the effectiveness of the proposed method.

Related papers

• TurboVGGT: Fast Visual Geometry Reconstruction with Adaptive Alternating Attention [21.29668311125256]
  TurboVGGT employs an efficient visual geometry transformer with adaptive alternating attention for fast multi-view 3D reconstruction.<n>In the adaptive sparse global attention, TurboVGGT adaptively learns representative tokens with varying sparsity levels for global geometry modeling.<n>Experiments on multiple 3D reconstruction benchmarks demonstrate that TurboVGGT achieves fast multi-view reconstruction while maintaining competitive reconstruction quality.
  arXiv  Detail & Related papers  (2026-05-14T03:24:09Z)
• AdaFuse: Accelerating Dynamic Adapter Inference via Token-Level Pre-Gating and Fused Kernel Optimization [84.25316984309725]
  Integration of dynamic, sparse structures with parameter-efficient adapters (e.g., LoRA) is a powerful technique for enhancing Large Language Models (LLMs)<n>Despite minimal increases in computational load, the computation latency often skyrockets, leading to decoding speeds slowing by over 2.5 times.<n>AdaFuse is a framework built on a tight co-design between the algorithm and the underlying hardware system to enable efficient dynamic adapter execution.
  arXiv  Detail & Related papers  (2026-03-12T12:46:42Z)
• Distill-then-Replace: Efficient Task-Specific Hybrid Attention Model Construction [3.9660062354591754]
  Transformer architectures deliver state-of-the-art accuracy via dense full-attention, but their quadratic time and memory complexity limits practical deployment.<n> Linear attention mechanisms offer linear or near-linear scaling yet often incur performance degradation.<n>We introduce a greedy layer replacement strategy that iteratively substitutes full attention blocks with linear ones while monitoring validation performance on the target task.<n>This yields a task-specific hybrid model in a single efficient pass, without costly re-training or neural architecture search, and can be applied to any pretrained full-attention backbone for diverse downstream tasks.
  arXiv  Detail & Related papers  (2026-01-16T02:01:40Z)
• iFlame: Interleaving Full and Linear Attention for Efficient Mesh Generation [49.8026360054331]
  iFlame is a novel transformer-based network architecture for mesh generation.<n>We propose an interleaving autoregressive mesh generation framework that combines the efficiency of linear attention with the expressive power of full attention mechanisms.<n>Our results indicate that the proposed interleaving framework effectively balances computational efficiency and generative performance.
  arXiv  Detail & Related papers  (2025-03-20T19:10:37Z)
• Systems and Algorithms for Convolutional Multi-Hybrid Language Models at Scale [68.6602625868888]
  We introduce convolutional multi-hybrid architectures, with a design grounded on two simple observations.<n>Operators in hybrid models can be tailored to token manipulation tasks such as in-context recall, multi-token recall, and compression.<n>We train end-to-end 1.2 to 2.9 times faster than optimized Transformers, and 1.1 to 1.4 times faster than previous generation hybrids.
  arXiv  Detail & Related papers  (2025-02-25T19:47:20Z)
• CARE Transformer: Mobile-Friendly Linear Visual Transformer via Decoupled Dual Interaction [77.8576094863446]
  We propose a new detextbfCoupled dutextbfAl-interactive lineatextbfR atttextbfEntion (CARE) mechanism. We first propose an asymmetrical feature decoupling strategy that asymmetrically decouples the learning process for local inductive bias and long-range dependencies. By adopting a decoupled learning way and fully exploiting complementarity across features, our method can achieve both high efficiency and accuracy.
  arXiv  Detail & Related papers  (2024-11-25T07:56:13Z)
• Linearly-evolved Transformer for Pan-sharpening [34.06189165260206]
  Vision transformer family has dominated the satellite pan-sharpening field driven by the global-wise spatial information modeling mechanism. Standard modeling rules within these promising pan-sharpening methods are to roughly stack the transformer variants in a cascaded manner. We propose an efficient linearly-evolved transformer variant and employ it to construct a lightweight pan-sharpening framework.
  arXiv  Detail & Related papers  (2024-04-19T11:38:34Z)
• Perceiving Longer Sequences With Bi-Directional Cross-Attention Transformers [13.480259378415505]
  BiXT scales linearly with input size in terms of computational cost and memory consumption. BiXT is inspired by the Perceiver architectures but replaces iterative attention with an efficient bi-directional cross-attention module. By combining efficiency with the generality and performance of a full Transformer architecture, BiXT can process longer sequences.
  arXiv  Detail & Related papers  (2024-02-19T13:38:15Z)
• CloudAttention: Efficient Multi-Scale Attention Scheme For 3D Point Cloud Learning [81.85951026033787]
  We set transformers in this work and incorporate them into a hierarchical framework for shape classification and part and scene segmentation. We also compute efficient and dynamic global cross attentions by leveraging sampling and grouping at each iteration. The proposed hierarchical model achieves state-of-the-art shape classification in mean accuracy and yields results on par with the previous segmentation methods.
  arXiv  Detail & Related papers  (2022-07-31T21:39:15Z)
• Combiner: Full Attention Transformer with Sparse Computation Cost [142.10203598824964]
  We propose Combiner, which provides full attention capability in each attention head while maintaining low computation complexity. We show that most sparse attention patterns used in existing sparse transformers are able to inspire the design of such factorization for full attention. An experimental evaluation on both autoregressive and bidirectional sequence tasks demonstrates the effectiveness of this approach.
  arXiv  Detail & Related papers  (2021-07-12T22:43:11Z)

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.