Command R+

As explained in Data Analysis with LLMs

Command R+ is a large language model developed by Cohere for generation, retrieval-augmented generation, and related language tasks.

Command R+ is a Cohere language model that can use external data to ground its replies. Cohere connectors can retrieve information from sources such as the web and databases, and Cohere shows which sources informed an answer so you can check them. section 8.2

Retrieval-augmented generation (RAG) adds information retrieved from an external source to the input used to generate an answer. This can help ground replies in data beyond the prompt. section 8.2.1 section 8.2

Compare answers with web search

In Cohere’s Playground, ask a factual question, then toggle the web connector on and off and compare the replies. When a reply uses external sources, follow its references to check whether they support the answer. section 8.2.1 section 8.2

Figure 8.2 Web interface for chatting with Cohere’s language model Command R+

Shows the Command R+ chat interface and where to enter a prompt and activate connectors. section 8.2.1

How Data Analysis with LLMs covers it

How deep it goes

Defines retrieval-augmented generation and describes connector use and source checking, without detailing a Python implementation.

Examples

Compares replies to factual questions with the web connector activated and deactivated, then checks cited sources.

What you can do after

Operate Command R+ in Cohere’s Playground by toggling web search and comparing answers to factual questions.

How it connects

Links Command R+ to Cohere connectors, web search, databases, and retrieval-augmented generation.

liveBookmanning.com

Read Data Analysis with LLMs on liveBook Get the book at manning.com

This explainer was generated by Manning's knowledge graph from the text of Data Analysis with LLMs. The section links point to the parts of the book it draws on.