Call for Papers
Background
The rapid growth of artificial intelligence has led to software systems that must make decisions in environments with many interacting components. These systems range from autonomous services and communication platforms to large scale digital infrastructures. In such settings, individual components often act with partial information and varying objectives. Understanding how these components learn, adapt and coordinate is an important challenge in both software engineering and machine learning. Research in areas such as reinforcement learning, game theory and collective intelligence has begun to provide tools that allow developers to study these interactions in a principled way. This symposium brings together researchers who are interested in the foundations and practice of learning in multi agent software environments.
Goal/Rationale
The central goal is to explore how software systems can learn to act effectively when their decisions depend on the behaviour of other agents. Traditional learning methods often assume a single learner or a stable environment, which does not reflect the interactive nature of many modern applications. Recent advances have introduced methods that allow multiple learners to adjust to one another, yet many open questions remain. These include how to manage conflicting objectives, how to ensure reliable learning in dynamic settings and how to design system level incentives that promote stable collective behaviour. By focusing on these challenges, the symposium aims to identify promising research ideas and practical solutions that can guide the development of future intelligent software systems.
Scope and Information for Participants
The symposium welcomes contributions that study learning and decision processes in systems with multiple interacting agents. Themes include coordination and competition in software environments, learning dynamics in strategic settings, interest alignment and conflict resolution in distributed systems, reinforcement learning with interacting learners, and methods that support transparency and reliability in collective decision making. We also encourage work that connects theoretical progress with software applications in areas such as autonomous services, communication platforms, digital markets and distributed control. Submissions may be conceptual, methodological or application focused. Participants will have the opportunity to discuss new ideas, present ongoing work and engage with others who are developing techniques for complex interactive systems.
