Skip to content

v0.2.0: Semantic Memory Search

Choose a tag to compare

View all tags
@dyascj dyascj released this 27 Jan 03:02
· 235 commits to main since this release
v0.2.0
7ee3d87
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

What's New

Semantic Memory Search

Your companion now finds memories by meaning, not just keywords. Ask about 'outdoor activities' and it will find memories about hiking - even without matching words.

How it works:

  • Uses Transformers.js with a local embedding model (~23MB, cached after first load)
  • Runs entirely in your browser - no API calls, no cost
  • Falls back to keyword search if model can't load

Technical Details

  • 384-dimensional embeddings stored in IndexedDB
  • Cosine similarity search blended with importance scores
  • Auto-backfills embeddings for existing facts on upgrade
  • Database schema updated to v3

Other Changes

  • Version number now injected from package.json at build time
  • Added CHANGELOG.md
Assets 2
Loading