Skip to content

Ising Model as Perception-Action Loop - Intrinsic Utility Comparison
Compare
Choose a tag to compare
@qianyangchen qianyangchen released this 17 Sep 07:54
· 7 commits to main since this release
v1.0.0
9d552de
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Learn about vigilant mode.

This release contains the full implementation of the Ising model as a perception-action loop, designed to compare four different intrinsic utility approaches:

  • Predictive information
  • Empowerment
  • Active inference (intrinsic component only)
  • Thermodynamic efficiency

The goal of this simulation is to identify the optimal regime in the parameter space for each utility measure using the Ising model as a common example.

Key features:

  • Jupyter notebook for running the simulations and visualising the comparisons, with full descriptions of the calculation process.
  • Required packages documented in requirements.txt
  • This release includes all the necessary scripts and notebooks for running the simulations and comparing the intrinsic utility measures. The codebase is designed to be extensible for further investigations in related fields.
Assets 2
Loading