For the most updated information on Active Blockference at the Institute, see:

active-blockference.activeinference.institute/

Active Blockference

Active Blockference is a project stewarded by participants in the Active Inference Institute. It consists of an open-source Python package combining cadCAD (complex systems modeling framework and library) and pymdp (an active inference Markov Decision Process engine) for the study of both theoretical and applied Active Inference.

Currently, Active Blockference is in active development and we welcome participants from all backgrounds to join our weekly discussions or contribute to the growing codebase at github.com/ActiveInferenceInstitute/ActiveBlockference.

Get involved

Join the Active Blockference channel in Active Inference Institute’s Discord server here.

Contributing to Coda

Most asynchronous collaboration happens in the Active Blockference Coda document . 

Participants are welcome to explore as well as to add comments, ideas, and projects for discussion in synchronous meetings.