International business machines corporation (20240185081). Learning Neuro-Symbolic World Models simplified abstract

Learning Neuro-Symbolic World Models

Organization Name

international business machines corporation

Inventor(s)

Don Joven Ravoy Agravante of Tokyo (JP)

Daiki Kimura of Midori-ku (JP)

Michiaki Tatsubori of Oiso (JP)

Learning Neuro-Symbolic World Models - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240185081 titled 'Learning Neuro-Symbolic World Models

Simplified Explanation

The patent application describes mechanisms for a model-based reinforcement learning system that includes a proprioception module to estimate the current state of an environment based on previous states and actions. The module also modifies the transition model to prevent past invalid actions from recurring in corresponding states.

• The proprioception module estimates the current state of an environment based on previous states and actions.
• The module modifies the transition model to prevent past invalid actions from recurring in corresponding states.

Potential Applications

The technology could be applied in autonomous vehicles, robotics, and gaming industries.

Problems Solved

1. Preventing past invalid actions from recurring in corresponding states. 2. Improving the accuracy of estimating the current state of an environment.

Benefits

1. Enhanced performance in model-based reinforcement learning systems. 2. Increased efficiency in decision-making processes for agents in dynamic environments.

Potential Commercial Applications

Optimizing resource allocation in supply chain management systems.

Possible Prior Art

There may be prior art related to reinforcement learning systems that focus on state estimation and action selection, but specific examples are not provided in the abstract.

How does the modified estimate of the transition model prevent past invalid actions from recurring in corresponding states?

The abstract does not provide detailed information on the specific mechanism used by the proprioception module to modify the transition model.

What are the specific industries or applications where this technology could have the most significant impact?

While the abstract mentions autonomous vehicles, robotics, and gaming industries as potential applications, it does not elaborate on the specific use cases or benefits in these sectors.

Original Abstract Submitted

mechanisms are provided for a model-based reinforcement learning (rl) computing system. a proprioception module receives a previous state of an environment and a previous action taken by an agent in the environment, and estimates a current state by using a transition model which receives a pair of state and action and produces a next state. the proprioception module modifies an estimate of the transition model so that the modified estimate of the transition model prevents a past invalid action from recurring in a corresponding state, where the past invalid action taken in the corresponding state is one that did not cause a change in state. the proprioception module passes the current state and the modified estimate of the transition model to a model-based rl computer model for generation of a next action to take in the environment.