Method and apparatus with autonomous driving decision learning

An electronic device comprising:

• one or more processors;
• and
• a memory storing instructions configured to cause the one or more processors to:
• receive a dataset related to driving from a database in which a driving scenario is stored;
• obtain a first output by inputting the dataset into a first large language model (LLM), wherein the first output comprises an inference of the first LLM based on the dataset;
• based on the first output, train a decision-trainer and a behavior-and-trajectory planner;
• generate a second LLM by updating the first LLM based on a result of the training, obtain a second LLM from the first LLM;
• and
• generate a third LLM by updating the second LLM based on the trained decision trainer and based on a selected traffic law.

A vehicle comprising:

• one or more processors;
• a memory storing instructions configured to cause the one or more processors to:
• display, by a display of the vehicle, a user interface for controlling the autonomous vehicle;
• execute a driving-decision algorithm of the vehicle;
• receive a dataset related to driving of the vehicle from a database in which a driving scenario is stored;
• obtain a first output by inputting the dataset into a first large language model (LLM), wherein the first output comprises an inference of the first LLM based on the dataset;
• based on the first output, train a decision trainer and a behavior-and-trajectory planner;
• generate a second LLM by updating the first LLM based on a result of the training;
• and
• generate a third LLM by updating the second LLM based on the trained decision-trainer and based on a selected traffic rule.