https://arxiv.org/abs/2401.09486

Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member
institutions, and all contributors. Donate
 
arxiv logo > cs > arXiv:2401.09486
[                    ]

Help | Advanced Search

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

  * Login
  * Help Pages
  * About

Computer Science > Machine Learning

arXiv:2401.09486 (cs)
[Submitted on 16 Jan 2024]

Title:LoMA: Lossless Compressed Memory Attention

Authors:Yumeng Wang, Zhenyang Xiao
Download a PDF of the paper titled LoMA: Lossless Compressed Memory
Attention, by Yumeng Wang and 1 other authors
Download PDF HTML (experimental)

    Abstract:The ability to handle long texts is one of the most
    important capabilities of Large Language Models (LLMs), but as
    the text length increases, the consumption of resources also
    increases dramatically. At present, reducing resource consumption
    by compressing the KV cache is a common approach. Although there
    are many existing compression methods, they share a common
    drawback: the compression is not lossless. That is, information
    is inevitably lost during the compression process. If the
    compression rate is high, the probability of losing important
    information increases dramatically. We propose a new method,
    Lossless Compressed Memory Attention (LoMA), which allows for
    lossless compression of information into special memory token KV
    pairs according to a set compression ratio. Our experiments have
    achieved remarkable results, demonstrating that LoMA can be
    efficiently trained and has very effective performance.

Subjects: Machine Learning (cs.LG); Computation and Language (cs.CL)
Cite as:  arXiv:2401.09486 [cs.LG]
          (or arXiv:2401.09486v1 [cs.LG] for this version)
          https://doi.org/10.48550/arXiv.2401.09486
          Focus to learn more
          arXiv-issued DOI via DataCite

Submission history

From: Yumeng Wang Dr. [view email]
[v1] Tue, 16 Jan 2024 09:18:46 UTC (930 KB)
Full-text links:

Access Paper:

    Download a PDF of the paper titled LoMA: Lossless Compressed
    Memory Attention, by Yumeng Wang and 1 other authors
  * Download PDF
  * HTML (experimental)
  * Other Formats

view license
Current browse context:
cs.LG
< prev   |   next >
new | recent | 2401
Change to browse by:
cs
cs.CL

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?)
[ ] 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

[ ] Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
[ ] DagsHub Toggle
DagsHub (What is DagsHub?)
[ ] 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?)
[ ] Connected Papers Toggle
Connected Papers (What is Connected Papers?)
[ ] 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