Everyone building GraphRAG tunes the same three dials: the retriever, the chunk size, the re-ranker....
Everyone building GraphRAG tunes the same three dials: the retriever, the chunk size, the re-ranker. Then the retrieved subgraph gets serialized into the prompt with json.dumps() — and nobody measures what that step costs.
I did. It costs more than everything else you tuned.
The benchmark
I compared 10 graph serialization formats — JSON, GraphML, RDF variants, edge lists, and others — on three axes:
Token count for the same property graph
I've spent the last several months building ANIMUS, an autonomous system in Rust that gives a local...
The chunk size forum thread is 47 replies long. Everyone's arguing about 512 vs 1024 tokens, hybrid...
How we measured a markdown knowledge graph as agent memory — lost to grep, rebuilt our search and...
The advice in 2026 is settled: chunk your documents, embed them, retrieve top-k, feed those to the...
In my last post I described graphlens — what it does, how it works — and along the way I casually...
If you've been exploring AI agents recently, chances are you've come across RAG (Retrieval-Augmented...
