Mario Sergio Rodriguez

Title: Mario Sergio Rodriguez and the Trust Layer for Large Language Models.

Introduction

Mario Sergio Rodriguez is an inventor based in San Francisco, California. His patent work includes a cloud-platform approach designed to add a configurable trust layer to interactions with large language models.

Latest Patents

Rodriguez is associated with 1 patent, titled “Trust Layer for Large Language Models.” The invention describes a cloud platform that receives a client’s prompt and applies the client’s configuration parameters to modify it before sending it to a large language model. After receiving the model’s response, the platform provides it to another model that estimates the likelihood that the response contains content from specified categories. The platform can then transmit the response, those probabilities, or both to the client.

Career Highlights

The patent highlights Rodriguez’s work on configurable controls for large language model interactions. Its described system combines prompt modification with content-category probability assessment, giving clients information that can help them evaluate model responses.

Collaborations

Rodriguez’s listed coworkers include Shashank Harinath and Eugene Wayne Becker. Their names are associated with his professional collaborations.

Conclusion

Mario Sergio Rodriguez’s patent describes a cloud-based trust layer that modifies prompts and assesses generated responses, reflecting an approach to configurable oversight of large language model outputs.