Verifiable State-Transition Memory

Validation

It was run inside an isolated container with no network access. This is the exact command and the real output it produced — captured process output, not written by a model.

$ python3 vstm.py
============================================================
 VSTM — Verifiable State-Transition Memory (Demo + Self-Test)
============================================================

[1] HMSM fitted — centroids:
    cluster 0: [0.35, 0.6, 0.15]
    cluster 1: [0.925, 0.05, 0.025]
    cluster 2: [0.1, 0.1, 0.85]
    labels: {'h1': 1, 'h2': 0, 'h3': 2, 'h4': 2, 'h5': 1, 'h6': 0}

[2] Processing live event stream...

  ACCEPTED: INIT -> IDLE (event=e1, rank=0.000, hash=ef32e9ff2f57...)
  ACCEPTED: IDLE -> INIT (event=e2, rank=0.000, hash=28b5de3e8035...)
  ACCEPTED: INIT -> IDLE (event=e3, rank=0.000, hash=e2c58c6543e7...)
  ACCEPTED: IDLE -> INIT (event=e4, rank=0.000, hash=05347001a874...)
  ACCEPTED: INIT -> IDLE (event=e6, rank=0.000, hash=b77c357568cd...)

[3] Rejected transitions (6):
  REJECTED: INIT -> IDLE (event=e5): invariant 'type-consistency' failed
  REJECTED: INIT -> ACTIVE (event=e5): invariant 'active-requires-min-confidence' failed
  REJECTED: INIT -> OVERLOAD (event=e5): invariant 'type-consistency' failed
  REJECTED: INIT -> ERROR (event=e5): invariant 'type-consistency' failed
  REJECTED: INIT -> COOLDOWN (event=e5): invariant 'type-consistency' failed
  REJECTE

A screenshot of that run.

A clean run proves this does what is shown above, in a CPU-only sandbox. It is a small research demo — not a production tool, and nothing here was published anywhere.