Large Language Model (LLM) and Generative Pre-trained Transformer (GPT) technology has the ability to generate responses to questions provided in text format. While such technology produces text content with plausible answers to a given question, it is difficult to judge the trustworthiness of the generated answers. It has been shown that answers generated by such technology may not always be factually accurate. This disclosure describes automated techniques to improve the credibility and confidence of machine-generated answers by using results from a search engine. Per the techniques, search results corresponding to a user input query are used to generate an augmented query that is provided as input to an LLM. Such augmentation adds context and can improve the quality and reliability of answers generated by the LLM. Further, the techniques can also be utilized to add citations to the generated answer and to indicate a confidence level in the generated answer.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.