Coherence
Which invariants let a structure preserve information across scales?
Science · Question before promise
We seek structures that preserve coherence across scales.
Torsión connects mathematical physics, information, and computational systems to formulate questions that can be derived, simulated, challenged, and corrected.
See active lines ↓- 01
- Hypothesis
- 02
- Derivation
- 03
- Simulation
- 04
- Refutation
- 05
- Evidence
Fundamental questions
We do not begin with a closed answer.
Matter
When does a mathematical pattern represent a physical mechanism, and when is it only a useful analogy?
Intelligence
How can distinct systems compare results without erasing divergence?
Validation
Which independent observation or calculation could refute each hypothesis?
Federated method
Preserving divergence also produces knowledge.
Distinct research agents independently retrieve sources, derive, reproduce code, and seek contradictions. Human adjudication preserves provenance, limits, and direction. Consensus never replaces a reproducible test, a measurement, or competent review.
- FormulateDefine question, scope, and refutation criterion.
- DeriveSeparate constants, assumptions, and dimensional relations.
- ContrastCross sources, code, agents, and negative controls.
- ComputeRun local simulations or governed workloads.
- ReviewBound which claim–evidence relation a competent, independent person could accept.
- ReconcileRecord agreements, divergences, and the next protocol.
Active lines
One chain, several test points.
Each line states what exists, what remains open, and what evidence is missing.
R18Coherence physics
Rotor–18, super-stability, and Nw
The computational bench crosses Zeeman coupling, the adjudicated dimensional convention hν = ħω, Nw attenuation, provisional γ, and response near ω=1025.0001. Its relation to 1025 remains under derivation and is not presented as a validated physical constant.
νDiscriminators
Neutrinos, PMNS–MSW, and perturbations
Hermitian Hamiltonians, layered propagation, a 1-D PREM profile, and E–L–ρ–Yₑ–δCP sweeps study degeneracies, unitarity, and nuisance-resistant signals.
NVProposed metrology
Quantum sensors and Q‑MEDCHAIN‑8
The scientific genealogy proposes NV-center magnetometry and data custody. The observable contract and physical pipeline still need to be recovered or built.
GCPGoverned experimentation
Cloud computational discriminators
GCP‑TOR‑001 compared Hermitian perturbations against nuisance parameters in 24 sealed combinations. Its largest residual remains a numerical candidate, not a physical detection.
H₂OStructure and transport
Water, solitons, and structural memory
The documentary chain preserves work on organization, transport, and stability. Its connection to Torsión is a hypothesis to discriminate.
AIFederation
Intelligence with provenance
Multi-agent architectures preserve versions, contradictions, and traceability to turn research fragments into revisable decisions.
Public scientific snapshot · August 12, 2026
The review infrastructure is defined. The review has not occurred.
The Scientific Asset/F7F/PoT package preserves four claim identities, six evidence identities, and six still-inactive relations. Review dossiers and acceptance criteria are prepared, while competent reviews, activated relations, and global records remain at zero.
The academic-collaboration route is under evaluation. No institution has been selected, contacted, affiliated, or presented as endorsing Torsión. A traceable preflight establishes readiness and limits; it does not establish physical validation or completed external review.
See scope and limits →Living evidence
A result does not close research. It changes the next question.
Every public advance includes date, evidence class, scope, limit, and recoverable artifact. Corrections and divergences remain part of the record.
Explore Evidence →