A network of agents is considered whose decision processes are described by the quantum decision theory previously advanced by the authors. Decision making is done by evaluating the utility of alternatives, their attractiveness, and the available information, whose combinations form the probabilities to choose a given alternative. As a result of the interplay between these three contributions, the process of choice between several alternatives is multimodal. The agents interact by exchanging information, which can take two forms: (i) information that an agent can directly receive from another agent and (ii) information collectively created by the members of the society. The information field common to all agents tends to smooth out sharp variations in the temporal behaviour of the probabilities and can even remove them. For agents with short-term memory, the probabilities often tend to their limiting values through strong oscillations and, for a range of parameters, these oscillations last for ever, representing an ever lasting hesitation of the decision makers. Switching on the information field makes the amplitude of the oscillations smaller and even can halt the everlasting oscillations forcing the probabilities to converge to fixed limits. The dynamic disjunction effect is described.
Collective, especially group-based, managerial decision making is crucial in organizations. Using an evolutionary theoretic approach to collective decision making, agent-based simulations were conducted to investigate how human collective decision making would be affected by the agents' diversity in problem understanding and/or behavior in discussion, as well as by their social network structure. Simulation results indicated that groups with consistent problem understanding tended to produce higher utility values of ideas and displayed better decision convergence, but only if there was no grou…
An open question in the study of emergent behaviour in multi-agent Bayesian systems is the relationship, if any, between individual and collective inference. In this paper we explore the correspondence between generative models that exist at two distinct scales, using spin glass models as a sandbox system to investigate this question. We show that the collective dynamics of a specific type of active inference agent is equivalent to sampling from the stationary distribution of a spin glass system. A collective of specifically-designed active inference agents can thus be described as implementin…
Shared and structured inputs undermine collective random choice by reasoning AI agents
Research Center for Advanced Science and Technology, The University of Tokyo（东京大学先端科学技术研究中心）；Department of Aeronautics and Astronautics, School of Engineering, The University of Tokyo（东京大学工学部航空宇宙工学科）
University of Innsbruck（因斯布鲁克大学）；Universität Konstanz（康斯坦茨大学）
Collective behavior, and swarm formation in particular, has been studied from several perspectives within a large variety of fields, ranging from biology to physics. In this work, we apply Projective Simulation to model each individual as an artificial learning agent that interacts with its neighbors and surroundings in order to make decisions and learn from them. Within a reinforcement learning framework, we discuss one-dimensional learning scenarios where agents need to get to food resources to be rewarded. We observe how different types of collective motion emerge depending on the distance …
Group size effects and collective misalignment in LLM multi-agent systems
Department of Mathematics, City St George’s University of London（伦敦城市圣乔治大学数学系）；IT University of Copenhagen（哥本哈根IT大学）；Pioneer Centre for AI, Copenhagen（哥本哈根先锋人工智能中心）；Departament de Física, Universitat Politècnica de Catalunya（加泰罗尼亚理工大学物理系）
Conformity Generates Collective Misalignment in AI Agents Societies
University of Konstanz（康斯坦茨大学）；Sony Computer Science Laboratories - Rome（索尼计算机科学实验室-罗马）；Sapienza University of Rome（罗马大学）；Max Planck Institute for Dynamics and Self-Organization（马克斯·普朗克动态与自组织研究所）；Institute for the Dynamics of Complex Systems, University of Gottingen（复杂系统动力学研究所，哥廷根大学）
When Is Collective Intelligence a Lottery? Multi-Agent Scaling Laws for Memetic Drift in LLMs
CBS–NTT Program in Physics of Intelligence（物理智能中心-NTT项目）；Center for Brain Science（脑科学中心）；Harvard University（哈佛大学）；Physics of Artificial Intelligence Group（人工智能物理研究组）；NTT Research, Inc.（NTT研究公司）
