Skip to content

Latest commit

 

History

History
 
 

README.md

Volume 2: Memory And Computation

This volume moves from individual assemblies to dynamics over time. The question changes from "what fired?" to "where does activity go next?"

It covers:

  • memorizing short stimulus sequences
  • inspecting recall with Long-Range Inhibition (LRI)
  • running deterministic finite-state utilities
  • sampling probabilistic automata

Notebooks

  • 01_sequence_memory_lri.ipynb: sequence memorization, overlap diagnostics, animated LRI recall, and winner-turnover inspection under one seeded parameter setting.
  • 02_fsm_and_pfa.ipynb: parity with FSMNetwork and independent samples from a small PFANetwork.
  • 03_lri_parameter_lab.ipynb: sweep refractory period and inhibition strength, then inspect one detailed LRI recall trace.

These notebooks demonstrate package utilities. The full theoretical results belong to the cited assembly-calculus literature.

Treat the outputs as traces from a small machine: trajectories, overlap matrices, and sample counts. If a sequence recall run is short or noisy, that is part of the lesson. The mechanism is visible enough to debug rather than hidden behind a single success/failure label.