Scientific research plays a crucial role in advancing human civilization, thanks to the efforts of a multitude of individual actors. Their behavior is largely driven by individual incentives, both explicit and implicit. In this paper, we propose and validate a multi-agent model to study the complex system of scholarly publishing and investigate the impact of incentives on research output. We use reinforcement learning to make the behavior of the actors optimizable, and guide their optimization with a reward signal that encodes the incentives. We consider various combinations of incentives and predefined behaviors and analyze their impact on both individual (h-index, impact factor) and overall indexes of research output. Our results suggest that, despite its simplicity, our model is able to capture the main dynamics of the system. Moreover, we find that (a) most incentives tend to favor productivity over quality and (b) incentives related to journal perceived reputation tend to result in waste of research efforts.

Multi-Agent Simulations are useful tools to predict the effects of public policies. In the last three years, with the concerns around the COVID-19 pandemic, several simulations were developed to understand the effects of lockdown, travel, etc. Even before that, MAS systems were used to plan disaster evacuation policies, transit policies, and many others. In this paper, we propose and analyze a mixed model that considers the effects of masking and large scale evacuations at the scale of a large university campus and its neighborhood. This project is part of a larger effort to create a simulator that considers how human mobility (pedestrian, public transportation, private transportation) interacts with large scale events (natural disasters, entrance examinations, pandemics) at a neighborhood level in the Japanese context. We evaluate how the simulator in its current state can reflect the effect of different masking policies on the spread of COVID-19 during an earthquake evacuation scenario.

Understanding spreading dynamics can help predict how a highly contagious disease can infect an entire population, how ideas propagate in societies, and how successful marketing campaigns emerge. In this study, we develop an agent-based model to highlight the role of individual heterogeneity in defining and shaping spreading dynamics. We select the case study of a virus spreading. The proposed model creates proximity networks in an urban environment, which is based on the city of Brussels. Various implementations of individual features and decision heterogeneity were examined. Our findings highlight the impact of individual irrationality and the size of social networks on emergent spreading and on the efficiency of local interventions.

We examine the effect of noise on societies of agents using an agent-based model of evolutionary norm emergence. Generally, we see that noisy societies are more selfish, smaller and discontent, and are caught in rounds of perpetual punishment preventing them from flourishing. Surprisingly, despite the detrimental effect of noise on the population, it does not seem to evolve away. In fact, in some cases it seems the level of noise increases. We carry out further analysis and provide reasons for why this might be the case. Furthermore, we claim that our framework that evolves the noise/ambiguity of norms is a new way to model the tight/loose framework of norms, suggesting that despite ambiguous norms’ detrimental effect on society, evolution does not favour clarity.

Recent minimal modeling work, following a dynamical approach to the phenomenology of body memory and the en-active approach to cognitive science, has served as a computational proof of concept in support of conceiving body memory as a relational property that arises from the history of interactions of a whole brain-body-environment system, rather than as contents within the brain. Particularly, some of these studies have been focused on investigating the minimal type of social memory, i.e., dyadic body memory, using the so-called embodied dyadic interaction models. Here, we expand the related work on dyadic body memory by employing a sample of the embodied dyadic interaction models, which has demonstrated, in line with previous related work in social interaction, that by evolving agent pairs to maximize their neural complexity, they consistently display mutually coordinated behavior, which cannot be possible to achieve in isolation. We aim to investigate the emergent behavioral patterns during the encounters between agents with “different” (i.e., because of proceeding from interactive or isolated primary environments) minimal social ontogenies. For this purpose, we propose a re-definition of the concept of social ontogeny as the shaping of “being social”, which involves body memory, as being arisen from shared histories of social interactions, and present three simulation experiments. Our results revealed the emergence of three core behavioral patterns: (1) mutually coordinated dyads, (2) “exaggerated-shy” dyads, and (3) limited-coordination dyads. An analysis of agents’ neural and behavioral complexity is also performed. We then draw loose analogies between our findings and real-world examples.

The progression of the global SARS-CoV-2 pandemic has been characterised by the emergence of novel ‘variants of concern’ (VOCs), which have altered transmission rates and immune escape capabilities. While numerous studies have used agent-based simulation to model the transmission and spread of the virus within populations, few have examined the impact of altered human behaviour in response to the evolution of the virus. Here we demonstrate a prototype simulation in which a simulated virus continually evolves as the agent population alters its behaviour in response to the perceived threat posed by the virus. Both mutations influencing intra-host and inter-host evolution are simulated. The model shows that evolution can dramatically reduce the effect of individual behaviour and policies on the spread of a pandemic. In particular only a small proportion of non-compliance with policies is sufficient to render countermeasures ineffective and lead to the spread of highly infectious variants.

Simulating the development of cities is interesting from the point of view of understanding human communities, but also brings benefits to city planners. However, understanding the expansion of land use and transportation networks is a known challenge. In this work, we investigate a city development model that combines a rule-based procedural road generation algorithm with multi-agent simulation of land choice and movement through the map using the generated land use and transportation network. The city map is organized in a grid, and an initial land value for each cell is calculated based on its geospatial features. Next, a set of agents are randomly initialized and perform actions on the city map, such as establishing residences, commuting, and trading in the exploration phase. Then, the actions performed by the agents are used as parameters for recalculating land prices and guiding the expansion of the road network, in a network development phase. We evaluate the emergence of geometrical patterns in the road network as well as land use and population distribution in the final map. We also compare maps generated using geographical data from selected locations to their corresponding real world settlements.