Humans and most other animals are known to be strongly driven by expected rewards or adverse consequences. The process of acquiring new skills or adjusting behaviors in response to positive outcomes is known as reinforcement learning (RL).

RL has been widely studied over the past decades and has even been adapted to train some computational models, such as some deep learning algorithms. Existing models of RL suggest that this type of learning is linked to dopaminergic pathways (i.e., neural pathways that respond to differences between expected and experienced outcomes).

Anne G. E. Collins, a researcher at University of California, Berkeley, recently developed a new model of RL specific to situations in which people’s choices have uncertain context-dependent outcomes, and they try to learn the actions that will lead to rewards. Her paper, published in Nature Human Behaviour, challenges the assumption that existing RL algorithms faithfully mirror psychological and neural mechanisms.

Over the past decades, roboticists have introduced a wide range of advanced systems that can move around in their surroundings and complete various tasks. Most of these robots can effectively collect images and other data in their surroundings, using computer vision algorithms to interpret it and plan their future actions.

In addition, many robots leverage large language models (LLMs) or other natural language processing (NLP) models to interpret instructions, make sense of what users are saying and answer them in specific languages. Despite their ability to both make sense of their surroundings and communicate with users, most robotic systems still struggle when tackling tasks that require them to touch, grasp and manipulate objects, or come in physical contact with people.

Researchers at Tongji University and State Key Laboratory of Intelligent Autonomous Systems recently developed a new framework designed to improve the process via which robots learn to physically interact with their surroundings.