# Robby is Not a Robber (anymore): On the Use of Institutions for Learning Normative Behavior

## TL;DR

This paper presents a framework for guiding reinforcement learning agents to follow human social norms, enabling transferability and abstraction of normative behaviors across different domains without relying on specific RL algorithms.

## Contribution

It introduces a method to encode social norms, guide learning with an automatic reward system, and transfer policies across domains, all while being algorithm-independent.

## Key findings

- Norms effectively guide RL agents toward normative behaviors
- The approach enables transfer of learned policies across different domains
- The method is compatible with various RL algorithms

## Abstract

Future robots should follow human social norms in order to be useful and accepted in human society. In this paper, we leverage already existing social knowledge in human societies by capturing it in our framework through the notion of social norms. We show how norms can be used to guide a reinforcement learning agent towards achieving normative behavior and apply the same set of norms over different domains. Thus, we are able to: (1) provide a way to intuitively encode social knowledge (through norms); (2) guide learning towards normative behaviors (through an automatic norm reward system); and (3) achieve a transfer of learning by abstracting policies; Finally, (4) the method is not dependent on a particular RL algorithm. We show how our approach can be seen as a means to achieve abstract representation and learn procedural knowledge based on the declarative semantics of norms and discuss possible implications of this in some areas of cognitive science.