rag
Latest in rag
Mistral AI explains how to score retrieval-augmented generation systems using an LLM as judge with structured outputs, covering retrieval and answer quality.
Anthropic’s Contextual Retrieval method adds context to document chunks before indexing, cutting failed retrievals by 49% and by 67% with reranking.
Google has released DataGemma, open models that use statistics from Google’s Data Commons, through RIG and RAG methods, to reduce LLM hallucination.
Contextual AI, founded by RAG pioneers Douwe Kiela and Amanpreet Singh, has unveiled RAG 2.0, which it says can improve RAG performance up to 10 times.
Infosys has integrated Meta’s Llama 3.1 into Infosys Topaz to process documents, video and audio, and uses it in a legal assistant built with RAG.
Rakuten Group has partnered with OpenAI to use AI across its ecosystem, drawing on its customer data to improve customer service, shopping and B2B consulting.
Microsoft’s guide to AI terms explains language models, grounding, RAG, orchestration and memory, transformer and diffusion models, and frontier models.