Amazon technologies, inc. (20240428787). GENERATING MODEL OUTPUT USING A KNOWLEDGE GRAPH
Contents
GENERATING MODEL OUTPUT USING A KNOWLEDGE GRAPH
Organization Name
Inventor(s)
Mahdi Namazifar of Oakland CA (US)
Devamanyu Hazarika of Sunnyvale CA (US)
Dilek Hakkani-tur of Los Altos CA (US)
GENERATING MODEL OUTPUT USING A KNOWLEDGE GRAPH
This abstract first appeared for US patent application 20240428787 titled 'GENERATING MODEL OUTPUT USING A KNOWLEDGE GRAPH
Original Abstract Submitted
techniques for constraining the results of a generative language model to valid information using knowledge-grounded documentation. a generative language model may generate invalid results, including compound entities and incorrect entity relations. the techniques include, for a given user inquiry, determining a set of documented information, from a particular knowledge base, that corresponds to the user inquiry. the techniques further include determining a subgraph from a knowledge graph representing the knowledge base, as well as determining a trie data structure representation of the set of documented information. the user inquiry and subgraph are provided as input to a trained generative language model for generating a response to the user inquiry. the techniques include using the trie data structure to validate that the generated response corresponds to real information from the set of documented information.
