Computing US 12,361,212 B2 Not in force

System and method for generating and extracting data from machine learning model outputs

The invention describes a system that improves the quality of responses from large language models (LLMs). It works by first ranking different LLMs, then picking the best ones to answer a question using prompts that can include customizable, structured parts. The system then combines their answers into a single, refined response and can learn from its past successes to get even better over time.

Why it matters: Filed in 2024, this system addresses the ongoing challenge of inconsistent LLM outputs. The increasing availability of diverse LLMs and the growing demand for highly reliable, structured responses make a robust refinement and orchestration system more valuable today.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedSoftware logic, machine learning models, LLM orchestration

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

+10