https://arxiv.org/abs/2502.07864

Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member
institutions, and all contributors. Donate
 
arxiv logo > cs > arXiv:2502.07864
[                    ]

Help | Advanced Search

[All fields        ]
Search
arXiv logo
Cornell University Logo
[                    ] GO
quick links

  * Login
  * Help Pages
  * About

Computer Science > Machine Learning

arXiv:2502.07864 (cs)
[Submitted on 11 Feb 2025 (v1), last revised 13 Feb 2025 (this
version, v2)]

Title:TransMLA: Multi-Head Latent Attention Is All You Need

Authors:Fanxu Meng, Zengwei Yao, Muhan Zhang
View a PDF of the paper titled TransMLA: Multi-Head Latent Attention
Is All You Need, by Fanxu Meng and 2 other authors
View PDF HTML (experimental)

    Abstract:Modern large language models (LLMs) often encounter
    communication bottlenecks on current hardware, rather than purely
    computational constraints. Multi-head Latent Attention (MLA)
    tackles this challenge by using low-rank matrices in the
    key-value (KV) layers, thereby allowing compressed latent KV
    states to be cached. This approach significantly reduces the KV
    cache size relative to traditional multi-head attention, leading
    to faster inference. Moreover, MLA employs an up-projection
    matrix to increase expressiveness, trading additional computation
    for reduced communication overhead. Although MLA has demonstrated
    efficiency and effectiveness in Deepseek V2/V3/R1, many major
    model providers still rely on Group Query Attention (GQA) and
    have not announced any plans to adopt MLA. In this paper, we show
    that GQA can always be represented by MLA while maintaining the
    same KV cache overhead, but the converse does not hold. To
    encourage broader use of MLA, we introduce TransMLA, a
    post-training method that converts widely used GQA-based
    pre-trained models (e.g., LLaMA, Qwen, Mixtral) into MLA-based
    models. After conversion, the model can undergo additional
    training to boost expressiveness without increasing the KV cache
    size. Furthermore, we plan to develop MLA-specific inference
    acceleration techniques to preserve low latency in transformed
    models, thus enabling more efficient distillation of Deepseek R1.

Comments: this https URL
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
Cite as:  arXiv:2502.07864 [cs.LG]
          (or arXiv:2502.07864v2 [cs.LG] for this version)
          https://doi.org/10.48550/arXiv.2502.07864
          Focus to learn more
          arXiv-issued DOI via DataCite

Submission history

From: Meng Fanxu [view email]
[v1] Tue, 11 Feb 2025 18:20:18 UTC (326 KB)
[v2] Thu, 13 Feb 2025 18:07:04 UTC (327 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled TransMLA: Multi-Head Latent
    Attention Is All You Need, by Fanxu Meng and 2 other authors
  * View PDF
  * HTML (experimental)
  * TeX Source
  * Other Formats

license icon view license
Current browse context:
cs.LG
< prev   |   next >
new | recent | 2025-02
Change to browse by:
cs
cs.AI

References & Citations

  * NASA ADS
  * Google Scholar
  * Semantic Scholar

a export BibTeX citation Loading...

BibTeX formatted citation

x
[loading...          ]
Data provided by:

Bookmark

BibSonomy logo Reddit logo
(*) Bibliographic Tools

Bibliographic and Citation Tools

[ ] Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
[ ] Connected Papers Toggle
Connected Papers (What is Connected Papers?)
[ ] Litmaps Toggle
Litmaps (What is Litmaps?)
[ ] scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
( ) Code, Data, Media

Code, Data and Media Associated with this Article

[ ] alphaXiv Toggle
alphaXiv (What is alphaXiv?)
[ ] Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
[ ] DagsHub Toggle
DagsHub (What is DagsHub?)
[ ] GotitPub Toggle
Gotit.pub (What is GotitPub?)
[ ] Huggingface Toggle
Hugging Face (What is Huggingface?)
[ ] Links to Code Toggle
Papers with Code (What is Papers with Code?)
[ ] ScienceCast Toggle
ScienceCast (What is ScienceCast?)
( ) Demos

Demos

[ ] Replicate Toggle
Replicate (What is Replicate?)
[ ] Spaces Toggle
Hugging Face Spaces (What is Spaces?)
[ ] Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
( ) Related Papers

Recommenders and Search Tools

[ ] Link to Influence Flower
Influence Flower (What are Influence Flowers?)
[ ] Core recommender toggle
CORE Recommender (What is CORE?)
[ ] IArxiv recommender toggle
IArxiv Recommender (What is IArxiv?)

  * Author
  * Venue
  * Institution
  * Topic

( ) About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and
share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have
embraced and accepted our values of openness, community, excellence,
and user data privacy. arXiv is committed to these values and only
works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community?
Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is
MathJax?)

  * About
  * Help

  * Click here to contact arXiv Contact
  * Click here to subscribe Subscribe

  * Copyright
  * Privacy Policy

  * Web Accessibility Assistance
  * arXiv Operational Status
    Get status notifications via email or slack