Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-19 14:13:35 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #826] [BUG]: Citations are wrong. #484

New Issue
Closed
opened 2026-02-22 18:19:44 -05:00 by yindo · 1 comment
yindo commented 2026-02-22 18:19:44 -05:00
Owner
Copy Link
Copy Source

Originally created by @sumitsodhi88 on GitHub (Feb 27, 2024).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/826

How are you running AnythingLLM?

Docker (local)

What happened?

The results given are correct but the citations given are completely wrong and unrealated.

Are there known steps to reproduce?

No response

Originally created by @sumitsodhi88 on GitHub (Feb 27, 2024). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/826 ### How are you running AnythingLLM? Docker (local) ### What happened? The results given are correct but the citations given are completely wrong and unrealated. ### Are there known steps to reproduce? _No response_
yindo added the possible bug label 2026-02-22 18:19:44 -05:00
yindo closed this issue 2026-02-22 18:19:44 -05:00
yindo commented 2026-02-22 18:19:44 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Feb 27, 2024):

Not a bug, this is just how vector databases work. They do semantic similarity which is not logic based, but instead math based (cosine similarity)

Some ways to improve citation results:

  • Use a larger dimension embedder model (OpenAI)
  • Increase the threshold for workspace (Settings > Vector Database)
  • Manually partition documents (aka make smaller documents for uploads) for more logical splitting as ANythingLLM will just take chunks of ~1000 chars
  • Use document pinning for critical documents so they are always in context

We are working on better retrieval methods beyond naive RAG

@timothycarambat commented on GitHub (Feb 27, 2024): Not a bug, this is just how vector databases work. They do semantic similarity which is not logic based, but instead math based (cosine similarity) Some ways to improve citation results: - Use a larger dimension embedder model (OpenAI) - Increase the threshold for workspace (Settings > Vector Database) - Manually partition documents (aka make smaller documents for uploads) for more logical splitting as ANythingLLM will just take chunks of ~1000 chars - Use document pinning for critical documents so they are always in context We are working on better retrieval methods beyond naive RAG
yindo changed title from [BUG]: Citations are wrong. to [GH-ISSUE #826] [BUG]: Citations are wrong. 2026-06-05 14:35:33 -04:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label possible bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#484
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?