Sony patents AI method that adjusts behavior weights in real time
Sony describes a method for training AI agents to handle multiple goals by combining different reward signals with adjustable weights. The system allows the AI to change these weights during operation to adapt its behavior, such as avoiding predicted negative events, without needing separate training for each specific task.
In the filing’s own words
Artificial intelligence (AI) policies are developed for artificial agents for various domains, where the behavior of such AI agents can be changed during run time by the agent adaptively adjusting weights for select parameters. A single policy can be trained to handle the selection of parameters across a predetermined range for each component of an artificial intelligent agent within a domain. The agents can be trained across a number of weights within the desired range for each component. These weights determine how much of a reward portion for each component should be considered by the agent during training. Thus, a formulation can be realized for UVFA-like goals based on compositional reward functions parameterized by their components' weights. The agents may be configured to predict a future negative event, where the agent can adaptively adjust the components' weights, in runtime, in efforts to avoid the predicted future negative event.
What exactly they claim (claim 1)
1. A method for operating an artificial intelligent agent based on a weighted composition of parametric reward functions, the method comprising: defining a reward function based on a state and an action as a linear combination of a plurality of parameterized reward functions and a weight for each parameter of the plurality of parameterized reward functions; sampling multiple dimensions of the weight and the parameter for each of the plurality of parameterized reward functions from either a continuous or a discrete distribution; training a single policy of the artificial intelligent agent over a continuous goal space including the plurality of parameterized reward functions represented by the continuous distribution of the weight and the parameter for each of the plurality of parameterized reward functions; and adaptively adjusting, during operation of the artificial intelligent agent, the weight for one or more selected parameters of the parameterized reward functions.
Facts
- Company
- Sony (SONY GROUP CORPORATION)
- Official title
- ADAPTIVE MODIFICATION OF POLICY BEHAVIOR THROUGH ADAPTIVELY CHANGING PARAMETERIZED REWARD WEIGHT UNIVERSAL VALUE FUNCTION APPROXIMATORS
- Publication number
- US20260268206A1
- Application number
- 19071277
- Filed
- Mar 5, 2025
- Published
- Sep 10, 2026
- Inventors
- 4
- Claims
- 20
Classified under
G06N 20/00 Physics & computingMore from Sony in AI & software
- Sony patents AI that plays video games in your style Oct 6, 2026
- Sony patents AI that swaps game visuals for accessibility formats Sep 17, 2026
- Sony patents AI that changes game characters based on your play Oct 1, 2026
- Sony patents method to turn 2D videos into high-detail 3D animations Oct 15, 2026
- Sony patents AI bots that mimic your friends' gaming styles Sep 8, 2026
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.