Systems and methods for normalizing responses from large ai models
AssigneeDealmaker Playbook, Inc.
InventorsMartin Hansen, Jon Stirling, Alana Dobbins
Life of this patent
Estimating the term…
Loading patent…
AssigneeDealmaker Playbook, Inc.
InventorsMartin Hansen, Jon Stirling, Alana Dobbins
Estimating the term…
The present disclosure describes systems and methods for using a generative pre-trained transformer (GPT) large language model (LLM) to generate an LLM completion response. In some aspects, an application is configured to run on a computing platform and perform communication with the GPT LLM based on a response generation module. Additionally, a database server may store a plurality of databases and is in communication with the application. In some cases, the application may be configured to receive an input comprising a solution type and a text from a user. The application may retrieve, from the database server, a set of prompt sequence instructions that provide instructions for assembling a solution input for the solution type. The GPT LLM takes the assembled solution input as input and generates an LLM completion response.
A system for using a generative pre-trained transformer (GPT) large language model (LLM) to generate an LLM completion response, comprising:
A method for using a generative pre-trained transformer (GPT) large language model (LLM) to generate an LLM completion response, comprising:
Dependencies are inferred from the claim text.