Skip to content

Conversation

yIwEN0713
Copy link

Description

Blog about OpenSearch v3 Neural Sparse Models.

Issues Resolved

Resolve #3912

Check List

  • Commits are signed per the DCO using --signoff

By submitting this pull request, I confirm that my contribution is made under the terms of the BSD-3-Clause License.

Signed-off-by: yIwEN0713 <[email protected]>
Copy link

Thank you for submitting a blog post!

The blog post review process is: Submit a PR -> (Optional) Peer review -> Doc review -> Editorial review -> Marketing review -> Published.

Copy link

Hi @yIwEN0713,

It looks like you're adding a new blog post but don't have an issue mentioned. Please link this PR to an open issue using one of these keywords in the PR description:

  • Closes #issue-number
  • Fixes #issue-number
  • Resolves #issue-number

If an issue hasn't been created yet, please create one and then link it to this PR.

Copy link
Member

@zhichao-aws zhichao-aws left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

pls take a ref at this PR to add your author information. And fix the render problem


## Behind the improvements: Technology highlights

### $L_0$-based sparsification
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

L_0 doesn't get rendered correctly. same below

Uploading image.png…


Similar to the tests described in our previous blog post, we evaluated the search relevance of the models on BEIR benchmark. The search relevance results are shown in the table below. All v3‑series models outperform their v2 and v1 counterparts, with v3‑gte achieving the highest relevance scores across all tests and setting a new record for OpenSearch neural sparse retrieval models. Notably, v3-gte sets a new state-of-the-art among publicly available sparse models.

---
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This table also doesn't get rendered correctly

@pajuric
Copy link

pajuric commented Aug 27, 2025

@natebower - Including you for technical review on this blog.

@natebower
Copy link
Collaborator

@natebower - Including you for technical review on this blog.

@pajuric It looks like this one may already be under tech review (also, I'm not able to provide tech reviews).

Copy link
Member

@zhichao-aws zhichao-aws left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. @pajuric I think we're good to move forward to the next review phase

@pajuric
Copy link

pajuric commented Aug 29, 2025

@natebower - Thanks for letting me know. We'll leave this one for Fanit when she returned early next week.

@kolchfa-aws

@kolchfa-aws kolchfa-aws self-assigned this Sep 2, 2025
Signed-off-by: Fanit Kolchina <[email protected]>
@kolchfa-aws
Copy link
Collaborator

Hey @yIwEN0713, doc review complete - nice job! Moving on to editorial review.

@kolchfa-aws kolchfa-aws added New blog New blog post Editorial review The blog is under editorial review labels Sep 8, 2025
@kolchfa-aws kolchfa-aws assigned natebower and unassigned kolchfa-aws Sep 8, 2025
Copy link
Collaborator

@natebower natebower left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Editorial review

Copy link
Collaborator

@natebower natebower left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks @yIwEN0713! LGTM

@pajuric This should be ready to publish.

@natebower natebower removed the New blog New blog post label Sep 8, 2025
@natebower natebower added Done and ready to publish The blog is approved and ready to publish and removed Editorial review The blog is under editorial review labels Sep 8, 2025
@natebower natebower assigned pajuric and unassigned natebower Sep 8, 2025

date: 2025-08-25
meta_keywords: search relevance, neural sparse search, OpenSearch semantic search
meta_description: Enhancing OpenSearch neural sparse models with GTE and LLM teachers.
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Updated meta:
meta_keywords: Neural sparse search, semantic search, OpenSearch sparse retrievers, search relevance, vector search, neural sparse models, GTE architecture, search relevance benchmarks, multilingual sparse retrieval, OpenSearch forum

meta_description: Discover OpenSearch v3's advanced neural sparse models and new multilingual retrieval capabilities, offering improved search relevance while maintaining efficient, inference-free performance

category:
- technical-posts

date: 2025-08-25
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Publish date: 2025-09-25

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Done and ready to publish The blog is approved and ready to publish
Projects
None yet
Development

Successfully merging this pull request may close these issues.

[BLOG] Neural sparse search v3 release
5 participants