TransMamba: Flexibly Switching between Transformer and Mamba

• URL: http://arxiv.org/abs/2503.24067v1
• Date: Mon, 31 Mar 2025 13:26:24 GMT
• Title: TransMamba: Flexibly Switching between Transformer and Mamba
• Authors: Yixing Li, Ruobing Xie, Zhen Yang, Xingwu Sun, Shuaipeng Li, Weidong Han, Zhanhui Kang, Yu Cheng, Chengzhong Xu, Di Wang, Jie Jiang,
• Abstract summary: This paper proposes TransMamba, a framework that unifies Transformer and Mamba.<n>We show that TransMamba achieves superior training efficiency and performance compared to baselines.
• Score: 43.20757187382281
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Transformers are the cornerstone of modern large language models, but their quadratic computational complexity limits efficiency in long-sequence processing. Recent advancements in Mamba, a state space model (SSM) with linear complexity, offer promising efficiency gains but suffer from unstable contextual learning and multitask generalization. This paper proposes TransMamba, a novel framework that unifies Transformer and Mamba through shared parameter matrices (e.g., QKV and CBx), and thus could dynamically switch between attention and SSM mechanisms at different token lengths and layers. We design the Memory converter to bridge Transformer and Mamba by converting attention outputs into SSM-compatible states, ensuring seamless information flow at TransPoints where the transformation happens. The TransPoint scheduling is also thoroughly explored for further improvements. We conducted extensive experiments demonstrating that TransMamba achieves superior training efficiency and performance compared to baselines, and validated the deeper consistency between Transformer and Mamba paradigms, offering a scalable solution for next-generation sequence modeling.

Related papers

• TransMamba: Fast Universal Architecture Adaption from Transformers to Mamba [88.31117598044725]
  We explore cross-architecture training to transfer the ready knowledge in existing Transformer models to alternative architecture Mamba, termed TransMamba.<n>Our approach employs a two-stage strategy to expedite training new Mamba models, ensuring effectiveness in across uni-modal and cross-modal tasks.<n>For cross-modal learning, we propose a cross-Mamba module that integrates language awareness into Mamba's visual features, enhancing the cross-modal interaction capabilities of Mamba architecture.
  arXiv  Detail & Related papers  (2025-02-21T01:22:01Z)
• MatIR: A Hybrid Mamba-Transformer Image Restoration Model [95.17418386046054]
  We propose a Mamba-Transformer hybrid image restoration model called MatIR.<n>MatIR cross-cycles the blocks of the Transformer layer and the Mamba layer to extract features.<n>In the Mamba module, we introduce the Image Inpainting State Space (IRSS) module, which traverses along four scan paths.
  arXiv  Detail & Related papers  (2025-01-30T14:55:40Z)
• Integrating Multi-Modal Input Token Mixer Into Mamba-Based Decision Models: Decision MetaMamba [0.0]
  Sequence modeling with State Space models (SSMs) has demonstrated performance surpassing that of Transformers in various tasks.<n>However, decision models based on Mamba, a state-of-the-art SSM, failed to achieve superior performance compared to enhanced Decision Transformers.<n>We propose the Decision MetaMamba (DMM), which augments Mamba with a token mixer in its input layer.
  arXiv  Detail & Related papers  (2024-08-20T03:35:28Z)
• An Empirical Study of Mamba-based Language Models [69.74383762508805]
  Selective state-space models (SSMs) like Mamba overcome some shortcomings of Transformers. We present a direct comparison between 8B-context Mamba, Mamba-2, and Transformer models trained on the same datasets. We find that the 8B Mamba-2-Hybrid exceeds the 8B Transformer on all 12 standard tasks.
  arXiv  Detail & Related papers  (2024-06-12T05:25:15Z)
• Mamba State-Space Models Are Lyapunov-Stable Learners [1.6385815610837167]
  Mamba state-space models (SSMs) were recently shown to outperform Transformer large language models (LLMs) across various tasks. We show that Mamba's recurrent dynamics are robust to small input changes. We also show that instruction tuning allows Mamba models to narrow this gap to 81% and Mamba-2 models to skyrocket over this gap to 132%.
  arXiv  Detail & Related papers  (2024-05-31T21:46:23Z)
• MambaTS: Improved Selective State Space Models for Long-term Time Series Forecasting [12.08746904573603]
  Mamba, based on selective state space models (SSMs), has emerged as a competitive alternative to Transformer. We propose four targeted improvements, leading to MambaTS. Experiments conducted on eight public datasets demonstrate that MambaTS achieves new state-of-the-art performance.
  arXiv  Detail & Related papers  (2024-05-26T05:50:17Z)
• MoEUT: Mixture-of-Experts Universal Transformers [75.96744719516813]
  Universal Transformers (UTs) have advantages over standard Transformers in learning compositional generalizations. Layer-sharing drastically reduces the parameter count compared to the non-shared model with the same dimensionality. No previous work has succeeded in proposing a shared-layer Transformer design that is competitive in parameter count-dominated tasks such as language modeling.
  arXiv  Detail & Related papers  (2024-05-25T03:24:32Z)
• Decision Mamba Architectures [1.4255659581428335]
  Decision Mamba architecture has shown to outperform Transformers across various task domains. We introduce two novel methods, Decision Mamba (DM) and Hierarchical Decision Mamba (HDM) We demonstrate the superiority of Mamba models over their Transformer counterparts in a majority of tasks.
  arXiv  Detail & Related papers  (2024-05-13T17:18:08Z)
• Is Mamba Effective for Time Series Forecasting? [30.85990093479062]
  We propose a Mamba-based model named Simple-Mamba (S-Mamba) for time series forecasting. Specifically, we tokenize the time points of each variate autonomously via a linear layer. Experiments on thirteen public datasets prove that S-Mamba maintains low computational overhead and achieves leading performance.
  arXiv  Detail & Related papers  (2024-03-17T08:50:44Z)
• Is Mamba Capable of In-Context Learning? [63.682741783013306]
  State of the art foundation models such as GPT-4 perform surprisingly well at in-context learning (ICL) This work provides empirical evidence that Mamba, a newly proposed state space model, has similar ICL capabilities.
  arXiv  Detail & Related papers  (2024-02-05T16:39:12Z)
• Mamba: Linear-Time Sequence Modeling with Selective State Spaces [31.985243136674146]
  Foundation models are almost universally based on the Transformer architecture and its core attention module. We identify that a key weakness of such models is their inability to perform content-based reasoning. We integrate these selective SSMs into a simplified end-to-end neural network architecture without attention or even blocks (Mamba) As a general sequence model backbone, Mamba achieves state-of-the-art performance across several modalities such as language, audio, and genomics.
  arXiv  Detail & Related papers  (2023-12-01T18:01:34Z)

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.