Skip to main content
TENET’s memory search combines lexical (BM25+) and semantic (embedding cosine similarity) scoring, merged via Reciprocal Rank Fusion.

​
Search Methods

MethodWhen UsedRequires API Key
bm25Keyword-focused queriesNo
embeddingSemantic/conceptual queriesYes
hybridBest of both (default)Partial — falls back to BM25

​
BM25+ Improvements

  • Stopword removal — cleaner term matching
  • Phrase detection — multi-word terms kept together
  • Adaptive b parameter — adjusts to corpus length variance
  • Positive IDF floor — common terms still contribute (BM25+ variant)
  • Query term weighting — rare terms get more weight

​
Boosts

BoostFactorCondition
Recency1.3xWithin 7 days
Decision1.4xType = decision
Feature1.2xType = feature

​
API

curl -X POST http://localhost:4360/api/memory/search \
  -H "Authorization: Bearer $TOKEN" \
  -d '{"query": "startup optimization", "limit": 5, "type": "decision"}'