Researchers, modelers, and engineers interested in multiagent systems, collective behavior, and Active Inference.
Multiagent modeling of collective behavior using Active Inference, with connections to ant colony research.
Lead: Daniel Friedman
This project is archived. It is no longer active, and any participation prompts below describe how it previously operated.
Active InferAnts is a multiagent modeling project that applies Active Inference to collective behavior, inspired by ant colony dynamics. It has produced a GitHub repository, a 2021 paper, and code developed within the Active Blockference project, with multiple realizations across different modeling contexts.
Active InferAnts develops multiagent models using Active Inference to simulate and understand collective behavior. The project draws inspiration from social insect biology — particularly ant colony foraging and organization — and applies generative modeling to understand how local agent behaviors produce collective intelligence without central coordination.
The project's foundational paper is Friedman, Tschantz, Ramstead, Friston, and Constant (2021), "An Active Inference Framework for Ant Colony Behavior," published in Frontiers in Behavioral Neuroscience — among the first Active Inference papers to model stigmergic (pheromone-mediated) communication. The project also maintains a GitHub repository with multiple code realizations and is developed in part through the Active Blockference project. The work spans individual agent models, collective foraging scenarios, and theoretical extensions.
Contributors with backgrounds in computational biology, multiagent systems, Active Inference, or behavioral ecology are welcome. Technical contributions (code, models) and conceptual work (theory, documentation) are both valued.
Related resources
Audience: Newcomer
Primary public community channel for conversation, coordination, and newcomer orientation.
Audience: Developer
Project repository for multiagent Active Inference modeling work.
Repositories
Audience: Developer
Python models and materials for ant-inspired multiagent Active Inference.
Audience: Developer
Public ants repository in the ActiveInferenceInstitute GitHub namespace.