/Patent Tracker
GooglePatent applicationAI & softwareInterest 7/10

Google patents method to teach AI agents when to use external tools

Patent drawing
Patent drawing from the official filing.

Google describes a system that helps artificial intelligence agents decide when to use outside tools to complete tasks. It uses a separate 'reward' model to check if the agent's choice to call a tool actually helps reach the goal, then feeds that feedback back to improve the agent's decision-making.

Why it mattersThis could make AI assistants better at handling complex, multi-step tasks by learning which external actions are actually useful rather than guessing.

In the filing’s own words

In one example aspect, the present disclosure provides an example computer-implemented method for generating feedback signals for training a machine-learned agent model. The example method can include obtaining an output of a machine-learned agent model, the output including a next state feature generated by the machine-learned agent model based on a sequence of preceding states. The example method can include processing, using a machine-learned reward model, the output and the sequence of preceding states to generate a quality indicator indicating a quality of the next state feature in view of the preceding states. The machine-learned reward model could be trained by retrieving reference data from a reference data source and computing one or more quality indicators in view of a respective training input and output(s), and the reference data. The example method can include outputting the quality indicator to a model trainer for updating the machine-learned agent model.

What exactly they claim (claim 1)

1. A computer-implemented method for training machine-learned agent models via reinforcement learning to improve tool utilization, the method comprising: obtaining, by a computing system comprising one or more processors, an output of a machine-learned agent model, wherein the output comprises an action feature generated by the machine-learned agent model based on a sequence of preceding states, the action feature specifying an action function call to a tool external to the machine-learned agent model; processing, by the computing system and using a machine-learned reward model, the output and the sequence of preceding states to generate a quality indicator that evaluates whether the action feature advances a goal associated with the sequence of preceding states; and providing, by the computing system, the quality indicator for input to a model trainer executing a reinforcement learning algorithm to update one or more parameters of the machine-learned agent model based on the quality indicator.

Facts

Company
Google (Google LLC)
Official title
Reinforcement Learning With Information Retrieval Feedback
Publication number
US20260311065A1
Application number
19711492
Filed
Jun 17, 2026
Published
Oct 8, 2026
Inventors
3
Claims
20

Classified under

G06N 20/00 Physics & computing
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Google in AI & software

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.