# n73 — pre-registration: does passivity own gravity's sign? (before code)

**The stakes.** The last door's one dial is Delta (bath detuning): the
n33 medium attracts in one branch, repels in the other. But a PHYSICAL
passive bath has modes at positive frequencies only — every Delta_m >
0 — so the induced Lamb coupling Omega = -sum g_m^2 Delta_m/(Delta_m^2
+ kappa^2/4) has a DEFINITE sign. The sign of gravity in this medium
is therefore OWNED by bath passivity (the same KMS structure that
makes Casimir attraction universal) — the question is only WHICH sign.
If the passive sign is the attractive-and-GR-paired branch: the wound
heals, nothing dialed, and the door stands with sign, pairing,
redshift and range all owned. If the passive sign is repulsion: the
last door dies by KMS, and the program closes with a complete map.

## Gate 0 (the discrepancy, resolved first)
n33's own text claims Delta = +1 attractive; the n71 reproduction
measured Delta = +1 repulsive (U[4] > U[8]). Run n33_bath.py VERBATIM
and read its sign line. Any divergence between its output and its
docstring is resolved in favor of the executed output, and recorded.

## Gate P (passivity)
Replace the single auxiliary mode per edge by a positive-frequency
multi-mode bath (Delta_m in {0.5, 1.0, 2.0, 4.0}, weights g_m^2 > 0,
same kappa): measure the sign of the mediated interaction U(x) and of
the clock pairing (phase slows near mass?) in the FULL model.
- P-LIFE: U attractive + phase slows: gravity's sign is a THEOREM of
  passivity in this medium. The door's last local wound closes.
- P-DEATH: U repulsive: monitored-bath gravity is repulsive for every
  physical bath; the door dies, and with it the framework's last
  gravitational sector. Recorded at full prominence either way.

## Honesty clauses
- The single-mode Delta < 0 branch corresponds to an INVERTED
  (population-inverted / active) bath — physically a gain medium. If
  P-DEATH fires, the honest epitaph is: "this gravity requires an
  active vacuum", and whether the real vacuum can be 'active' in the
  monitored sector becomes the final question rather than a settled
  kill — declared now to prevent post-hoc spin in either direction.
- First moments, N = 12, scalar sector; TT junction untouched.

## RESULTS (same day) — GATE 0 RESOLVED, P-DEATH FIRES: HEADSTONE 26
- Gate 0: n33's own executed output reads "Delta=+1 ... REPULSIVE" —
  the n71 reproduction was right, the docstring's remembered claim was
  wrong. The attractive branch was always the INVERTED bath.
- Gate P: the passive multi-mode bath (Delta_m = 0.5, 1, 2, 4, all
  positive; induced Omega = -0.0038, definite sign as passivity
  demands) gives REPULSION between like masses, measured in the full
  model. MONITORED-BATH GRAVITY IS REPULSIVE FOR EVERY PHYSICAL
  PASSIVE BATH. The sign is owned — by the wrong branch.
- HEADSTONE 26: the last door, as a passive-vacuum theory, is dead.
  What the epitaph leaves standing is exactly what the prereg declared
  before the verdict: attraction in this medium requires an ACTIVE
  vacuum — a monitored sector holding population inversion, a gain
  medium, a vacuum with a metabolism. That is no longer a dial; it is
  a physical hypothesis with consequences (KMS violation in one
  sector; an energy source — noting, without weight, that the
  framework's own Lambda-as-idle-metabolism language always described
  the monitor as a consumer). THE FINAL QUESTION OF THE PROGRAM, and
  it is a physics question, not a knob: IS THE VACUUM ACTIVE IN ITS
  MONITORED SECTOR? Everything else on the gravity side is now closed
  by a measurement or a theorem with its name attached.
- The quantum sectors are untouched by this stone: the electrostatic
  reading (n69), the knee (n48), and the redshift-is-quantum result
  (n72) all live in sectors that do not depend on the sign of Omega.

## n74 (same day): THE ACTIVE VACUUM STANDS
The hypothesis is not a knob — it is measurement thermodynamics: a
vacuum that audits continuously is a working measuring device, and a
demon needs a battery; the monitored sector of a measuring vacuum
cannot be KMS-passive. The inverted mode (Delta < 0) is exactly the
effective oscillator of a gain medium. Measured (n74_active_vacuum.py):
- A1 PASS: the inverted monitored medium is dynamically STABLE (full
  spectrum in the left half-plane up to the declared regularization).
  No runaway: the active vacuum is a consistent dynamical object.
- A2 PASS: the complete weak-field trio survives pumping — attraction,
  GR sign pairing, and the quantum-phase redshift (2.32% tracking) all
  hold in the active medium.
- A3 HONEST ZERO: the pump's power bill is invisible at Gaussian order
  (NESS carries exactly vacuum noise; P ~ 0 to regularization). The
  metabolism cost — and with it the Lambda-take-2 candidate — lives in
  the microphysics that maintains the inversion, beyond the linearized
  model. Open, named, not claimed.

## THE PROGRAM'S CLOSING MAP (2026-07-09, end of the two-day audit)
Gravity, in this framework, costs exactly three things, each now
measured or named: (1) a LOCAL potential field (the monitored medium
has it; drainage never did — the dichotomy); (2) PLANCK'S CONSTANT
(redshift is quantum phase — classical rates see only tides, theorem
at 0.00%); (3) an ACTIVE vacuum (passivity owns the repulsive sign —
attraction is the signature of a vacuum that measures). Standing
sectors: emergent electrostatics (n69, 5/5), the decoherence knee
(n48), thirteen predictions, twenty-six headstones, one formula, one
manuscript awaiting its author's read — and three open frontiers:
the pump's microphysics (Lambda), the tensor junction (beta = 1, the
quadrupole), and units. Every closed door has a measurement's name on
it. That is the account.