Method and apparatus for using a generative large language model for audiological fitting and device control

A method for fitting a hearing device comprising:

• receiving, by a processor via a user interface of the hearing device or a computing device, a first signal indicative of a natural language input, wherein the natural language input is indicative of an issue experienced by a user with the hearing device or the computing device;
• determining, by the processor, a first embedding based on the natural language input, wherein the first embedding is related to a first dimension of multiple dimensions;
• receiving, by the processor, a second signal indicative of:
• a setting of the hearing device or the computing device, or
• a metric measured by the hearing device or the computing device;
• determining, by the processor, a second embedding based on the setting or the metric, wherein the second embedding is related to a second dimension of multiple dimensions;
• generating, by the processor based on the first embedding and the second embedding, a first embedding vector, wherein the first embedding vector is indicative of values associated with each of the first embedding related to the first dimension and the second embedding related to the second dimension;
• determining, by the processor, a closest match between the first embedding vector and an embedding vector of an embedding vector of a plurality of embedding vectors, wherein each of the plurality of embedding vectors is associated with one of a plurality of related solutions for possible hearing issues with the hearing device or the computing device;
• and
• sending, by the processor to the user interface of the hearing device or the computing device, a third signal indicative of a reply output, the reply output comprising information related to a solution of the related solutions for the issue experienced by the user.

A fitting apparatus for a hearing device comprising:

• a processor;
• and
• a memory having stored thereon:
• non-transitory computer readable instructions executable by the processor, and
• a plurality of embedding vectors and related solutions for possible hearing issues with the hearing device or a computing device,
• and further wherein, upon execution of the instructions, the fitting apparatus is configured to:
• receive, via a user interface of the hearing device or the computing device, a first signal indicative of a natural language input, wherein the natural language input is indicative of an issue experienced by a user with the hearing device or the computing device;
• determine an embedding based on the natural language input;
• send, after receiving the first signal indicative of the natural language input, a second signal indicative of a follow-up request to the user interface of the hearing device or the computing device, wherein the follow-up request requests more information related to the issue experienced by the user;
• and
• receive, via the user interface of the hearing device or the computing device, a third signal indicative of a follow-up answer to the follow-up request,
• generate, based on the embedding and at least in part based on the follow-up request or the follow-up answer, a first embedding vector;
• determine a closest match between the first embedding vector and an embedding vector of the plurality of embedding vectors;
• and
• send, to the user interface of the hearing device or the computing device, a fourth signal indicative of a reply output, the reply output comprising information related to a solution of the related solutions for the issue experienced by the user.