Semantic clustering for machine self-learning

This invention helps a machine learning system learn from its mistakes with natural language. It takes user inputs, converts them into numerical codes using a large language model, and then groups similar codes together. If some inputs in a group cause errors, the system looks at information from other, successfully processed inputs in the same group to figure out how to fix the problematic ones or improve its understanding. The claims specifically focus on using an LLM to create these codes and identifying errors based on natural language understanding metrics.

Why it matters: Filed as large language models (LLMs) were rapidly advancing, the capabilities and efficiency of LLM encoders have significantly improved since 2023. This makes the core process of generating embeddings and assessing natural language understanding (NLU) errors more robust and practical to implement now.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedSoftware logic, LLM integration, NLU metrics

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.

This one

+10