Articles tagged: Latency Optimization

Anchor Detection for RAG: Parallel Detectors, Then One LLM Call at the End

A technique for efficient RAG that uses lightweight parallel detectors to identify semantic anchors before making a single, targeted LLM c...