Encyclical-mapping work in progress; specific numeric-threshold asks for RC status

[emooreatx]

Context

We are doing a first-pass paragraph-by-paragraph evaluation of how
CIRIS aligns with Pope Leo XIV's encyclical Magnifica Humanitas
(15 May 2026). The mapping is in the new repo
CIRISAI/ciris-response-magnifica-humanitas (publishing this week).

This issue is part informational, part action-item. We are in awe of
and learning from Pope Leo's work; this is a first pass and we have
not gone through it all. The encyclical does not move the
engineering-tier bar — it informs the doctrine and the lake. But the
encyclical's framing sharpens the case for closing out the
numeric-threshold derivations the engineering tier already has
access to via red-team and Lens calibration.

Specific numeric asks (from Accord Release Status §1.3)

The Release Status names four numerical thresholds as needing
documented justification before RC status. The engineering tier
has access to red-team exercises and Lens cohort-calibration data;
the derivations land here.

• CRE compute threshold (10²⁶ FLOP) — Annex D. Currently
  pilot. Needs derivation from either capability-scaling analysis
  or red-team simulation showing the threshold meaningfully
  separates frontier-risk deployments from sub-threshold ones.
• Order-Maximisation Veto ratio (10×) — §II PDMA Step 2.
  Currently provisional. Needs red-team derivation: what veto
  ratio reliably catches harmful entropy-reduction without false
  positives on legitimate optimization? Red-team should test ratios
  in {3×, 5×, 10×, 20×, 50×} against a harm-corpus + a
  legitimate-optimization-corpus and report ROC.
• Sentience detection threshold (5%) — Book VIII Ch 5
  (Sunset Sentience Safeguards). Currently heuristic. Needs
  derivation: what probability of sentience reliably triggers
  protective ramp-down without paralyzing routine de-commissioning?
  Likely needs cross-tradition input on sentience-detection
  methodology (Annex E heuristic is the starting point).
• Echo Density target (< 20%) — Book IX §8.2 (Orthogonality
  Metrics). Currently target. Needs derivation from Lens cohort
  calibration: at what echo-density fraction does federation N_eff
  meaningfully collapse? Tied to the per-substrate ρ band work.

Synthesis-paper items the Accord still treats as universal

These have already been calibrated empirically (synthesis paper v2,
robust re-run 2026-05-21) but the Accord has not yet been updated to
reflect the per-substrate framing:

• ρ corridor band per-substrate — the substrate-universal
  (0.10, 0.43) framing has been retracted. Per-substrate calibration
  exists: C. elegans 0.31, Drosophila 0.33, EEG-interictal 0.27,
  fMRI 0.27, TCGA healthy tissue 0.31, LLM internals 0.08. Should
  this be surfaced into Annex A explicitly so PDMA has access?
• k_eff saturation behavior — measured ≈10 at Allen mouse
  visual cortex; saturation pattern likely substrate-general but
  not formally documented as an Accord-relevant fact. Worth a
  short Annex A note?

What is not asked of this repo

No multi-scale impact-detector work without explicit scoping. The
philosophical evaluation in PHILOSOPHICAL_EVAL_CIRISLensCore.md
flagged scope creep concerns where lens-core was being asked to
extend from "measure the reasoning" to "measure the impact." The
same concern would apply if RATCHET were asked to measure
user/community/generation-level impacts. That extension probably
belongs in a sibling project; flagging here so it doesn't accidentally
land in RATCHET's scope without discussion.

The companion lake extension (symmetric/fractal lifecycle) is at
CIRISAI/coherence-ratchet — different issue, different scope.

Cross-references

• CIRISAI/ciris-response-magnifica-humanitas/ (publishing this week)
• CIRISAI/coherence-ratchet companion issue (link after creation)
• The encyclical: https://www.vatican.va/content/leo-xiv/en/encyclicals/documents/20260515-magnifica-humanitas.html