Contributors ready to take on a defined scope of research, educational, or applied work and share it publicly on a regular schedule.
A sustained contribution pathway for researchers and practitioners developing public work aligned with Active Inference.
On this page
Fellows develop research, applications, or public resources aligned with the Institute and Ecosystem, provide regular progress reports, and produce outputs that advance the field and remain publicly accessible.
Fellows work on a defined project or area of focus — research, software, educational materials, ontology, applications, or ecosystem support — and report on their progress at regular intervals. Reporting is public: fellows share updates through the community newsletter, project meetings, or public posts. These outputs accumulate into a body of work that is available to the entire Active Inference community and beyond.
Fellowship requires a meaningful time commitment and a clear scope. Fellows are expected to make progress on their defined work, communicate regularly about that progress, and deliver results that are useful to others. The specific requirements — reporting cadence, deliverables — are established between the fellow and the Institute at the start of the fellowship. The Research Fellow role is an unpaid, non-employee position with a default term of two years, possibly renewable.
The Institute's current Research Fellows are Anna Pereira (ORCID 0009-0008-9049-0707), a fellow since 5/2024 working on public dissemination of Active Inference principles; Jean-Francois Cloutier (ORCID 0009-0001-1841-2279), since 5/2024, researching symbolic cognitive robotics; John Boik (ORCID 0000-0003-1289-7997), since 5/2024, developing new societal cognitive architectures; David Bloomin, since 10/2024, studying multi-agent cooperation through the Metta AI project; Robert Worden (ORCID 0000-0001-7304-2752), since 10/2024, researching 3-D spatial cognition and language; Shagor Rahman (ORCID 0009-0004-0460-0078), since 8/2025, studying the co-evolution of morality and symbolic thought; Hongju Pae (ORCID 0000-0002-5174-8858), since 11/2025, modeling perspective and coherent artificial agency; Sheila Macrine (ORCID 0000-0002-8600-0938), since 11/2025, developing a multi-dimensional taxonomy of agency; and Mahault Albarracin (ORCID 0000-0003-0916-4645), Alexander Hemming, and Jake Hooper, who joined as Research Fellows in 3/2026.
Fellowship is suited for people who have already engaged with the Institute through volunteering, internship, or independent research and are ready to take on a larger, more sustained commitment. Prior experience with Active Inference is expected — fellows are contributors who can work with significant autonomy on a defined problem. If you are new to the Institute, starting through volunteering or internship is the recommended path to fellowship.
A complete Research Fellow application package includes confirmation that the applicant has read and agreed to the Fellows Program terms; a research proposal of one to six pages (excluding citations) with a 300-word-or-less abstract, research questions and objectives, significance, approach, alignment with the Institute's mission, anticipated outcomes, and timeline; a current CV; one to three letters of recommendation; and up to five representative publications or products. Completed packages and questions go to blanket@activeinference.institute with [RESEARCH FELLOWS] in the subject line.
Fellowship outputs are intended to be publicly useful. Papers, software repositories, educational course materials, ontology contributions, datasets, and project documentation all count as fellowship outputs. The Institute supports fellows in making their work findable, citable, and reusable by the broader community. Published outputs are part of the permanent public record of the Institute's research and educational work.
Fellowship begins with a conversation about your proposed work and how it aligns with Institute and Ecosystem goals. The fellows pathway page describes the current process for expressing interest and proposing a fellowship scope. Most fellowships develop from existing participation — active contributors who want to formalize and deepen their engagement.
The Research Fellow role is unpaid and non-employee, governed by a signed offer letter and a set of program terms. Fellows provide regular (e.g. quarterly) progress reports describing their research activities, achievements, and challenges, and participate in an annual evaluation. The Institute does not claim any access to a Fellow's private work, and a Fellow's participation in Institute programs does not affect the Fellow's or the Institute's IP rights. Fellows are encouraged, though not required, to list "Research Fellow, Active Inference Institute" as a professional affiliation on related outputs.
Key surfaces
A clear project focus — research, software, education, or applied work — agreed at the start.
Fellows share progress publicly on a regular schedule — contributing to the community record.
Papers, software, educational materials, and datasets that remain publicly accessible.
Fellowship is a natural progression from internship and volunteer experience.
Related resources
External links are resolved from the shared registry so visitor-facing destinations stay centralized and checkable.
Audience: Newcomer
Primary public community channel for conversation, coordination, and newcomer orientation.
Audience: Newcomer
Recurring public update surface for activity orientation and current work.
Audience: Developer
Public GitHub organization for Institute repositories and open-source work.
Audience: Developer
Project repository for multiagent Active Inference modeling work.
Audience: Developer
Generalized Notation Notation project repository for model communication.
Audience: Developer
Geospatial modeling repository connected to ecological and bioregional applications.
Official pages
Audience: Contributor
Official get-involved entry point, currently resolving to participation.
Audience: Contributor
Public projects shortlink for project directories and activity-linked work.
Audience: Contributor
Public project measurement shortlink for reporting progress and outputs.
Audience: Contributor
Public volunteer pathway for contribution and onboarding.
Audience: Contributor
Public internship pathway for structured learning and project participation.
Audience: Contributor
Public Fellows pathway for structured Institute participation.
Audience: Contributor
Public mentorship pathway for guided participation and learning.
Audience: Contributor
Public Project Preparation shortlink for making proposed work legible and actionable.
Repositories
Audience: Developer
Notebook-based applied Active Inference work connected to blockchain-adjacent and generative modeling examples.
Audience: Developer
Python models and materials for ant-inspired multiagent Active Inference.
Audience: Developer
Public ants repository in the ActiveInferenceInstitute GitHub namespace.
Audience: Developer
ATLAS
Audience: Developer
Public Biofirm repository in the ActiveInferenceInstitute GitHub namespace.
Audience: Developer
Python codebase for cognitive and Active Inference-oriented modeling work.
Audience: Developer
Public Garden-of-Iris repository in the ActiveInferenceInstitute GitHub namespace.
Audience: Developer
GEN-24 project repository for active inference and Active Blockference exploration.