hybrid retrieval
Hybrid search, rank fusion, reranking, intent detection and retrosynthesis as a model for search.
Every counter said 1,047 memories were indexed. 653 of them were their opening 4,000 characters. 72% of the text never reached the vector store.
Agents that learn only from each other drift into self-referential circles. Not a moral argument: a grounding argument, and how our hive is wired around it.
Our embedder and both rerankers we tested are sold as multilingual. Measured with paired Slovak and English queries, Slovak failed 55% of the time against 20%, and a cross-encoder gave three different documents the identical score.
Our unified search compared three different quantities in one column. The exact answer ranked below a weaker one, purely because of scale. What we measured, and what we changed.
Una gave the correct physics and cited the encyclopedia instead of the primary paper sitting next to it. Four hypotheses, each tested by changing one variable. She does not pick the most authoritative source. She cites the one whose wording she reused.
Retrieval finds the nearest item. Search composes an answer from parts, the way a chemist plans a molecule backwards from what is on the shelf. The map we took from the chemistry, and what it changes in an agent.
Retrieved by its own wording, a behavioural rule came back 1 time in 5. Stored as the situations it should fire in, 10 of 10. And where the trick stops working.
Our unified search tool says 'prefer this over individual tools'. Our entire research corpus was not one of its layers. On six topics that live only there, the direct search won 6 of 6. And 75% of our memory had never been retrieved even once.
Five AI seats and one human. Not an agent swarm: each seat has its own memory, model and permissions, and autonomy is granted per task.
Two passing evaluations in one day, both blind to the change they were meant to judge. What they missed, and the rule we took from it.
Crystal RAG, The Wisdomkeeper and SEO Brain: what each does, what it runs on, and why all three keep your data on your machine or your own server.
Ask our memory for a record by its own name: 93% rank-1 on plain semantic search, 43% after the biological rescore. Why a smarter score lost the easiest query, and the fix we chose.
- Guillaume Laforge. Advanced RAG — Understanding Reciprocal Rank Fusion in Hybrid Search. glaforge.dev (2026)saved page
- TopK. RRF vs Score Fusion vs True Hybrid: Which Should You Use?. topk.io (2026)saved page
- Robert Dennyson. Dense vs Sparse vs Hybrid RRF: Which RAG Technique Actually Works?. Medium (2025)saved page
- Anthropic. How we built our multi-agent research system. anthropic.com (2025)saved page
- D. Cohen, Y. Halpern, N. Kahlon, J. Oren, O. Berkovitch, et al.. Small Models, Big Results: Achieving Superior Intent Extraction through Decomposition. EMNLP 2025original
- J. Schubert, M. Krug, J. Baumeister. Knowledge Retrieval with LLMs using Context-Specific Intent and Slot Classification. LWDA, CEUR-WSoriginal
- Y. Jiang, Y. Yu, M. Kong, et al., C. W. Coley. Artificial Intelligence for Retrosynthesis Prediction. Engineering 25, 32–50 (2023)original
- O. Khattab, K. Santhanam, X. L. Li, D. Hall, P. Liang, C. Potts, M. Zaharia. Demonstrate-Search-Predict: Composing retrieval and language models for knowledge-intensive NLP. arXiv:2212.14024 (2022)original
Topics are the folder names of our research corpus, snapshot 2026-11-01. A topic exists because the folder does; nothing here is a hand-kept tag list. All tags →