Canonical KS core
The executed numerical core is the favored Hamiltonian artifact. The printed v2.4 equations must be rewritten to match it.
→Canonical theory v2.4 · findings baseline Revision 5
From quantum
transition to a
measured limit.
Aeternus is a developing loop-quantum-cosmology framework studying a finite transition inside an effective Kantowski–Sachs geometry. The latest audit preserves the geometric attractor while finding that the tested background is not FLRW-like.
Revision 5 findings · 19 August 2026 · Paper I reframing in progress
EFFECTIVE KS GEOMETRY FINITE TRANSITION
TESTED ATTRACTOR NOT FLRW-LIKE
CURRENT SCIENTIFIC POSITION · REVISION 5
The attractor survives.
The cosmological identity does not.
The numerical core supports a finite transition followed by an exact, anisotropic Kantowski–Sachs attractor. The executed core is now the favored Hamiltonian artifact, while the printed v2.4 equations require correction.
Across the tested local neighborhood, radial expansion persists, the angular two-sphere freezes, and shear approaches σ/Havg ≈ √3. The directional classifier returns NOT_FLRW_LIKE. This is a bounded local negative result—not a no-go theorem for every KS–LQC branch.
ANIMATED CONCEPTUAL MODEL
Collapse. Transition. Tested outcome.
The motion illustrates the current causal sequence. It is conceptual—not a numerical spacetime simulation or a derived Penrose diagram.
01Space is squeezedKS anisotropic contraction
02Collapse stopsfinite transition · θ = 0
04Tested outcomelocal · finite-window result
inside the effective geometrytested sequence →measured limit
Theory v2.4 · Revision 5 findings · conceptual, not to scale · local and finite-window result
Supported core: a finite transition followed by a frozen-sphere KS attractor.
Manuscript repair: rewrite the printed Hamiltonian and correct the e-fold clock.
Measured limit: the tested background does not sustain an FLRW-like phase; no CMB connection is claimed.
THE REVISION 5 CASCADE
What the calculations support—and where they stop
Each link carries its present evidentiary status. The sequence now ends at the tested directional result rather than an observational handoff.
Finite transition
The tested effective dynamics replace the classical singular limit with a finite transition surface.
→Frozen-sphere attractor
After the transition, radial expansion continues while the angular two-sphere freezes and shear remains order unity.
→Directional viability test
Four high-precision configurations and two transient configurations return NOT_FLRW_LIKE in the tested neighborhood.
→Paper I reframing
The active result is the KS attractor and its negative cosmological viability test—not a parent-mass or CMB prediction.
CURRENT CONSOLIDATED RESULT
A locally robust non-FLRW attractor
Four high-precision directional anchors and two additional transient configurations support the negative verdict in the tested neighborhood.
σ/HavgMeasured
1.732040–1.732043Four high-precision configurations converge near √3, showing persistent order-unity anisotropy.
H₂Measured
≈ 0The angular two-sphere freezes while the preferred radial direction continues expanding.
G2Blocked
NOT_FLRW_LIKEThe tested frozen-sphere neighborhood fails the directional horizon and mode-freezing criteria.
δNKSRevised
12.83 ± 0.1Corrected net endpoint displacement—not elapsed duration—after restoring the volume-clock factor.
NEGATIVE COSMOLOGICAL RESULT
The tested frozen-sphere neighborhood does not provide an inflationary background for our universe. Sphere-direction horizons grow and modes fail to freeze. The result establishes local robustness only; it does not rule out every branch or an exit at arbitrarily late time.
AUDIT & EVIDENCE STATUS
What changed in Revision 5
The update separates audited mathematical structure, required corrections, bounded negative evidence and unfinished reproducibility work.
Hamiltonian audit
The code integrates the audited numerical core, but the manuscript's printed Hamiltonian and H₁ rate do not match. Rewrite Sections 2.6–3.2 from the extracted core.
E-fold clock
The implementation omitted the p_c^(1/6) contribution. Correct a_eff, requote every e-fold-dependent value, and switch to N only after the final turning point.
Directional viability
The negative result is locally robust and finite-window bounded. It is not a theorem about the full KS–LQC branch or arbitrarily late time.
Reproducibility
The step-cap comparison is bit-identical, but the formal checkpointed G2b continuation driver remains incomplete.
RESEARCH & DEVELOPMENT ROADMAP
Progress is earned,
not presumed.
Aeternus is a long-term theoretical-physics research project. Its central hypotheses are not assumed to be correct. Each future stage depends on the model surviving the stage before it.
Publication of the implementation case study, future software repair and acceptance, and scientific validation of Aeternus physics are three distinct milestones. None establishes the others.
Historical Current Future or conditional
01
Initial formulation
The conceptual framework, Kantowski–Sachs and LQC calculations, numerical models, and early manuscript material established the hypotheses that would later enter systematic verification.
Research baseline—not independently established results.02
Model-to-model re-evaluation
Earlier claims were reconstructed from their mathematics and software. The audit exposed a manuscript–Hamiltonian mismatch, an incorrect e-fold clock, persistent directional anisotropy, and limits in onset and gradient-flow claims.
Affected conclusions were corrected, narrowed, superseded, or withdrawn.03
Verification pipeline
Machine-checkable claim, evidence, falsification, retraction, promotion, provenance, and manuscript-authorization controls replaced informal scientific recordkeeping.
AT-CLM-0004 historically passed G8 separate-model cold review and entered VERIFIED, an internal pipeline lifecycle state—not scientific verification or empirical confirmation.04
Implementation case study published
The Zenodo technical report documents implementation, testing, failure discovery, and evolution of the claim-management system. The case study preserves failures rather than reporting an accepted software release.
Publication does not establish software acceptance, architectural novelty, general effectiveness, or validation of Aeternus physics.05
Repair and evaluate the pipeline
The evaluated Revision 8 snapshot retains A-9 FAIL 2/4 and package verification FAIL 42. Repair defects, rerun the required checks, and document a separate acceptance decision before claiming an accepted release.
Revision 8: REVIEW EVIDENCE ONLY — NOT A CANDIDATE RELEASE.06
Canonical reconstruction
The project is migrating surviving results into the integrity system, reproducing core calculations, formalizing assumptions, rebuilding affected simulations, and mapping parameter dependence and stability.
Priority: a smaller set of well-supported results.07
Build a defensible model
Complete claim migration; formalize the core model; test mathematical consistency; recover required limits of established physics; establish novelty against the literature; and validate a robust numerical implementation.
Advancement is conditional on each preceding stage surviving review.08
Prediction and confrontation
Derive a quantitative prediction that differs from credible alternatives, compare it with existing observations, and release enough equations, code, parameters, data, and history for external reproduction.
Observations may support, constrain, eliminate, or falsify a version of the theory.09
Manuscript and peer review
Release manuscripts only when included claims meet their evidentiary requirements, then submit mature work for independent scientific peer review and incorporate material criticism into the correction process.
This future physics-manuscript milestone is separate from the published implementation case study and from empirical validation.10
Scientific validation of Aeternus physics
Test distinctive predictions through observation or experiment. Confirmation, constraints, null results, and falsification would all provide scientifically useful information.
Aeternus advances only as far as the evidence allows.Conjecture→Formal model→Established limits→Distinctive prediction→Independent test
SCIENTIFIC INTEGRITY PIPELINE
Claims advance only when
the evidence advances.
Aeternus does not treat a statement as verified because it appears in a manuscript, simulation output, AI response, changelog, or earlier discussion. Models may propose and challenge claims; only the governed integrity pipeline may promote them.
Canonical records track internal claim lifecycle and authorization. VERIFIED is an internal pipeline lifecycle state, not scientific verification, empirical confirmation, or external peer review.
CORE RULENo scientific claim should become more certain merely because it has been repeated.
01→
Claim proposed
A precise, falsifiable proposition enters as PROPOSED or TESTING—not as a fact.
02→
Evidence registered
Derivations, data, simulations, sources, and reviews are linked to identifiable artifacts and hashed where required.
03→
Dependencies checked
Assumptions, conventions, software, parameters, sources, and upstream claims are made explicit.
04→
Verification gates
The claim must pass the analytical, numerical, convergence, literature, or dependency checks appropriate to it.
05→
Model-to-model re-derivation
A separate AI model works from the specification under input-separation controls. This is not independent human replication or peer review.
06→
Separate-model cold review
Another AI model reviews the claim with prior verdicts and framing withheld where appropriate. This is not external peer review or independent scientific review.
07→
Validator check
Executable rules reject broken records, invalid transitions, missing evidence, and unauthorized actions.
08→
Canonical promotion
An authorized action can move a qualifying claim to VERIFIED, an internal lifecycle state—not scientific verification or empirical confirmation.
09✓
Publication decision
Manuscript eligibility is authorized separately; VERIFIED does not automatically mean publishable.
Claims do not have to reach VERIFIED.
BLOCKEDSTALESUPERSEDEDRETRACTED
If a requirement fails, evidence changes, or an upstream dependency is invalidated, the affected status changes and remains part of the historical record.
TWELVE OVERLAPPING SAFEGUARDS
Safeguards and their limits.
Definition integrityClaims must be precise enough to test or invalidate.
Evidence integritySupporting material is identifiable and linked.
Artifact integrityGate evidence is hashed to preserve identity.
Provenance integrityExact results, approximations, assumptions, and observations remain distinct.
Dependency integrityUpstream changes propagate to dependent claims.
Verification integrityEach claim type must satisfy the right scientific checks.
Input separation integritySeparate-model re-derivation is intended to avoid reuse of the generating analysis; model errors may still be shared.
Review integritySeparate-model cold-review controls aim to limit framing effects; they do not constitute external peer review.
Software integritySchemas, validators, and adversarial tests aim to enforce the procedure; Revision 8 retains recorded failures.
Governance integrityAI systems cannot promote their own conclusions.
Publication integrityScientific status and manuscript authorization stay separate.
History integrityCorrections, supersessions, and retractions remain traceable.
IMPORTANT LIMIT
The pipeline is intended to protect the record. It does not prove the theory.
A validator can enforce process, provenance, evidence linkage, dependency, lifecycle, and authorization rules. Scientific correctness must still come from mathematics, reproducible computation, established physics, peer-reviewed literature, independent scrutiny, and ultimately observation or experiment.
RESEARCH OUTPUTS
Research Outputs / Publications
Technical report · Implementation case study
Verification-Gated Claim Management for AI-Assisted Theoretical Research: An Aeternus Implementation Case Study
This technical report documents the implementation, testing, failure discovery, and evolution of Aeternus’s verification-gated claim-management system for AI-assisted theoretical research. It is a research-integrity and reproducibility case study; it does not validate Aeternus Theory or claim that the architecture is novel or generally effective.
DOI: 10.5281/zenodo.22836334PAPER I SCOPE RESET
The observational bridge is removed from the active claim set.
Parent-mass inversion, Nref, CMB templates, polarization asymmetry and other observational constructions are archived pending a viable background. The current paper should center on the numerically studied KS–LQC attractor, persistent shear, directional horizon failure and the resulting negative cosmological viability test.
ACTIVEHamiltonian core · finite transition · anisotropic attractor · local G2 verdict
REVISION 5 CRITICAL PATH
Repair the manuscript before extending the model
The next phase is implementation: correct the volume clock, requote e-fold-dependent values, rewrite the manuscript Hamiltonian from the audited numerical core, and preserve a machine-readable provenance ledger.
Discuss or review the work ↗C1Correct the volume clockUse a_eff = (p_b√p_c)^(1/3), add the regression identity, and requote every e-fold result.
G3Apply the retraction registerReplace 13.69 ± 0.1 with 12.83 ± 0.1 as net displacement and repair the printed equations.
G4Complete reproducibilityFinish the checkpointed G2b continuation driver with explicit validity gates and structured provenance.
G7Map open boundariesTreat the p_c0 = 1.60 branch loss and φ = 34 root boundary as a separate research campaign.