[2604.20920] Simplified Sparse Attention via Gist Tokens
-->
Computer Science > Machine Learning
arXiv:2604.20920 (cs)
[Submitted on 22 Apr 2026 (v1), last revised 26 Jun 2026 (this version, v2)]
Title:Simplified Sparse Attention via Gist Tokens
Authors:Yuzhen Mao, Michael Y. Li, Emily B. Fox<br>View a PDF of the paper titled Simplified Sparse Attention via Gist Tokens, by Yuzhen Mao and 2 other authors
View PDF<br>HTML (experimental)
Abstract:Sparse attention can reduce the cost of long-context inference, but most variants introduce new architectural components. We introduce Simplified Sparse Attention (SSA), a simpler approach to sparse attention that requires no architectural changes. Concretely, we first perform continued pretraining on sequences interleaved with gist tokens. We optimize the standard next-token loss as usual, but the gist tokens use an attention mask to restrict what parts of the context the language model can attend to; this teaches the model to pack each chunk's important information into the gist tokens. At inference time, SSA scores chunks via attention between the current query and the small set of gist tokens, selectively unfolding the top-k chunks by reintroducing their corresponding raw tokens. Since the query is scored only against the gist tokens, we avoid the memory-bandwidth cost associated with naive scoring against the full KV cache, without requiring the auxiliary KV cache approach used by sparse attention methods. On LongBench, SSA consistently outperforms compression and inference-time sparse-attention baselines under the same compression ratio. More strikingly, in retrieval-augmented generation, SSA can even outperform full attention after continued pretraining by over 5.7 points. We attribute this to the ability of SSA's selective unfolding, which concentrates attention on the query-relevant chunks and effectively filters out noise. SSA further extends to a hierarchical gist-of-gist variant (H-SSA) that achieves log-linear decoding complexity while maintaining or improving accuracy at high compression ratios up to 32x. The code is available at this https URL.
Subjects:
Machine Learning (cs.LG)
Cite as:<br>arXiv:2604.20920 [cs.LG]
(or<br>arXiv:2604.20920v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2604.20920
Focus to learn more
arXiv-issued DOI via DataCite
Submission history<br>From: Yuzhen Mao [view email]<br>[v1]<br>Wed, 22 Apr 2026 04:22:32 UTC (157 KB)
[v2]<br>Fri, 26 Jun 2026 06:02:50 UTC (97 KB)
Full-text links:<br>Access Paper:
View a PDF of the paper titled Simplified Sparse Attention via Gist Tokens, by Yuzhen Mao and 2 other authors<br>View PDF<br>HTML (experimental)<br>TeX Source
view license
Current browse context:
cs.LG
next >
new<br>recent<br>| 2026-04
Change to browse by:
cs
References & Citations
NASA ADS<br>Google Scholar
Semantic Scholar
export BibTeX citation<br>Loading...
BibTeX formatted citation
×
loading...
Data provided by:
Bookmark
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?)
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<br>(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? |<br>Disable MathJax (What is MathJax?)