Retrieval-Augmented Generation (RAG)

GPTKB entity

Predicate	Object
gptkbp:instanceOf	machine learning technique natural language processing method
gptkbp:architecture	retriever-generator
gptkbp:canBe	gptkb:BART gptkb:Chroma gptkb:OpenAI_GPT gptkb:LangChain gptkb:LlamaIndex gptkb:Qdrant gptkb:Weaviate gptkb:T5 gptkb:Elasticsearch gptkb:FAISS gptkb:Milvus gptkb:OpenSearch gptkb:Pinecone gptkb:Google_BERT gptkb:RAG-Sequence gptkb:RAG-Token
gptkbp:category	generative model AI architecture retrieval-based model
gptkbp:citation	gptkb:NeurIPS_2020 gptkb:Lewis_et_al.,_2020 gptkb:Retrieval-Augmented_Generation_for_Knowledge-Intensive_NLP_Tasks
gptkbp:combines	generator retrieval
gptkbp:enables	open-domain question answering knowledge-intensive tasks
gptkbp:firstPublished	2020
gptkbp:generation	large language model
gptkbp:generatorComponent	gptkb:model sequence-to-sequence model
https://www.w3.org/2000/01/rdf-schema#label	Retrieval-Augmented Generation (RAG)
gptkbp:improves	factual accuracy
gptkbp:input	query
gptkbp:openSource	gptkb:Haystack gptkb:Hugging_Face_Transformers
gptkbp:output	generated text
gptkbp:proposedBy	gptkb:Facebook_AI_Research
gptkbp:reduces	hallucination
gptkbp:relatedTo	gptkb:knowledge_representation LLM retrieval-augmented language model
gptkbp:retrieverComponent	dense retriever
gptkbp:retrieverType	vector search
gptkbp:usedIn	question answering chatbots document summarization
gptkbp:uses	external documents retrieval corpus
gptkbp:bfsParent	gptkb:Patrick_Lewis
gptkbp:bfsLayer	5