Prodigii’s Large Behavioral Models are built on Active Inference. That is a checkable claim rather than a marketing one, because Active Inference is a published research program with two decades of peer-reviewed literature behind it. Dr. Karl Friston, whose work established it, advises Prodigii on the scientific foundations of Active Inference. The papers below are his and his co-authors’ published research; Prodigii’s implementation is its own, and none of this literature is an endorsement of it.
This page is the reading list. Every reference below was verified against its publisher record — author list, journal, year, and DOI. Nothing here is Prodigii’s own work, and nothing here should be read as a description of Prodigii’s product. It is the science the product rests on, so that an engineer or an investor can evaluate the foundation without asking anyone for permission.
Read in this order if you are starting cold: the tutorial first for a working grasp, then the process theory for the formal account, then the book if you want the whole program.
Start here
Smith, R., Friston, K. J., & Whyte, C. J. (2022). A step-by-step tutorial on active inference and its application to empirical data. Journal of Mathematical Psychology, 107, 102632.doi.org/10.1016/j.jmp.2021.102632
The most practical entry point. It builds the machinery incrementally and works through applying it to real data, which is the difference between recognizing the terminology and being able to use it.
Parr, T., Pezzulo, G., & Friston, K. J. (2022). Active Inference: The Free Energy Principle in Mind, Brain, and Behavior. The MIT Press.doi.org/10.7551/mitpress/12441.001.0001
The book-length treatment, and open access from the publisher. If you want one source that covers the whole program rather than one result, this is it.
The formal account
Friston, K., FitzGerald, T., Rigoli, F., Schwartenbeck, P., & Pezzulo, G. (2017). Active Inference: A Process Theory. Neural Computation, 29(1), 1–49.doi.org/10.1162/NECO_a_00912
The paper that states Active Inference as a process theory: how a system that maintains a model of its world selects among the actions available to it. This is the reference for what “evaluates alternatives” means formally.
Da Costa, L., Parr, T., Sajid, N., Veselic, S., Neacsu, V., & Friston, K. (2020). Active inference on discrete state-spaces: A synthesis. Journal of Mathematical Psychology, 99, 102447.doi.org/10.1016/j.jmp.2020.102447· open preprint: arXiv:2001.07203
The synthesis for discrete state spaces — the setting enterprise decision problems usually live in. The closest thing in the literature to an implementer’s reference.
Friston, K. (2010). The free-energy principle: a unified brain theory?Nature Reviews Neuroscience, 11(2), 127–138.doi.org/10.1038/nrn2787
The origin paper for the principle Active Inference is derived from. Historically important and widely cited; not the place to start, but the place the rest comes from.
Comparison and context
Sajid, N., Ball, P. J., Parr, T., & Friston, K. J. (2021). Active Inference: Demystified and Compared. Neural Computation, 33(3), 674–712.doi.org/10.1162/neco_a_01357· open preprint: arXiv:1909.10863
Sets Active Inference against reinforcement learning directly. The most useful single paper for anyone whose first question is how this differs from the methods they already know.
Friston, K. J., Ramstead, M. J. D., Kiefer, A. B., et al. (2024). Designing ecosystems of intelligence from first principles. Collective Intelligence, 3(1).doi.org/10.1177/26339137231222481· open preprint: arXiv:2212.01354
Extends the principles from a single agent to networks of them — shared objectives, shared beliefs, coordinated action. This is the part of the literature closest to what an orchestration layer has to do.