Social Reward

Social reward research explores how receiving positive feedback from others influences behavior and learning, aiming to understand both its underlying mechanisms and its implications for various systems. Current investigations focus on modeling internal representations of social reward, developing algorithms to incentivize cooperation in multi-agent systems (e.g., through reward sharing and decomposition), and leveraging large-scale datasets of human social interactions to improve AI models' alignment with community preferences. This work has significant implications for advancing artificial intelligence, particularly in areas like human-robot interaction and the design of more socially beneficial autonomous systems, as well as for understanding fundamental aspects of human social behavior.