Method and device for learning a vehicle interior conversational agent

Abstract

The present invention relates to a method and a device for learning a vehicle interior conversational agent, the conversational agent implementing a large language model, called LLM, the conversational agent being configured to implement communication in natural language between a user of the vehicle and at least one on-board system of the vehicle. For this purpose, deep learning of the LLM based on human preferences is implemented with optimization of the LLM policy by projecting updated parameters of a current policy into a finite space of policies to obtain a new LLM policy. Figure for abstract: Figure 1