# Frontier experiment atlas v1 **Date:** 2026-10-26 **Coverage:** 20/20 frontier areas; 55/55 maintained topics; 39/39 detector bundles. This atlas converts every canonical frontier into a realistic controlled experiment. “Experiment” is used in the broad scientific sense: a laboratory intervention where a physical carrier exists, a reproducible computation where the object is discrete, and a theorem/falsification benchmark where uniform limits or an action are the actual missing datum. A blocked fundamental frontier is not disguised as build-ready hardware. ## Portfolio | ID | Frontier | Mode | Readiness | Indicative scale | |---|---|---|---|---| | EX-001 | Structural reachability | computational | ready_now | 2–4 weeks; one proof engineer plus one independent reproducer | | EX-002 | Representation and normalization stability | computational | ready_now | 4–8 weeks; proof engineer and data engineer | | EX-003 | Detector resolution and semantic fibres | computational_hybrid | ready_now | 6–10 weeks; modelling, statistics and domain-review trio | | EX-004 | Positivity, rank and response cones | hybrid | ready_after_design | 3–6 months on an existing microwave or transport platform | | EX-005 | Orientation and phase lifts | laboratory | ready_after_design | 6–12 months using an integrated-photonic or microwave platform | | EX-006 | Projector holonomy and coherent transport | laboratory | ready_after_design | 9–18 months including foundry cycle | | EX-007 | Dynamical selection and attraction | laboratory | ready_after_design | 9–18 months; can share EX-006 hardware | | EX-008 | Measurement backaction and conditioning | laboratory | ready_now_on_existing_hardware | 6–12 months on a superconducting qutrit or continuously monitored qubit | | EX-009 | Thermodynamic cost and resource budgets | laboratory | ready_now | 4–8 months on an optical-feedback trap | | EX-010 | Metastability, rare events and hysteresis | laboratory | ready_now | 6–12 months; long unattended acquisition required | | EX-011 | Spectral boundary dynamics | laboratory | ready_now | 4–9 months with microwave or acoustic resonators | | EX-012 | Material response and fluctuation forces | laboratory | multi_stage | 18–36 months across materials, transport, optics and force metrology teams | | EX-013 | Multichannel transport and hydrodynamic rank | laboratory_computational | ready_after_design | 12–24 months including microfabrication | | EX-014 | Topology, defects and protected transport | laboratory | ready_now | 6–12 months with acoustic or topoelectric resonator arrays | | EX-015 | Disorder, causal spread and scale | laboratory_computational | ready_after_design | 6–12 months on a programmable resonator or qubit array | | EX-016 | Arithmetic and code islands | exact_computational | ready_now | 2–6 months; algebra, coding and proof expertise | | EX-017 | Operator and carrier translation | hybrid | staged | 12–24 months after a source-native map is written | | EX-018 | Consistency, scattering and continuum control | theorem_computation | ready_now | 6–18 months depending on whether a theorem survives | | EX-019 | Flavour, gravity and cosmology | conditional_hybrid | conditional_on_action | 12–36 months after a frozen action and matching packet exist | | EX-020 | Exceptional and unification interpretation | theory_computation | conditional_on_model | 12–36 months for one fully specified candidate | | EX-021 | Flavour, gravity and cosmology | theory_kernel_then_public_data_recast | source_complete_amplitude_required | 2–6 months for the amplitude audit; 3–9 months for a public-likelihood recast after release | | EX-022 | Spectral boundary dynamics | laboratory_atomic_spectroscopy | ready_on_established_platform | 6–12 months on an equipped trapped-ion platform; helium transfer arm is a separate specialist campaign | ## Experiment cards ### EX-001 — Pinned-kernel reachability tournament **Frontier:** FA-01 — Structural reachability (constraint sieve) **Mode/readiness:** computational; ready_now **Indicative scale:** 2–4 weeks; one proof engineer plus one independent reproducer **Topics:** OT-033, OT-035 **Detector bundles:** DET-001, DET-002 **Platform.** A pinned Lean toolchain and a stratified sample of real declarations, synthetic edge cases and deliberately unreachable constructor trees. **Hypothesis.** Kernel elaboration and constructor closure identify a strict reachable image whose failures are stable under harmless presentation changes. **Interventions** - Vary constructor depth, arity and closure while holding the target type fixed - Blind the analyst to reachable versus adversarial cases - Repeat under pinned transparency and compact-constructor policies **Recorded observables** - Elaboration outcome and inferred type - Minimal failed constraint - Reachability certificate or counterexample - Runtime and memory **Controls** - Alpha renaming - Literal expansion - Known overflow witnesses - Tree Euler and closure identities **Analysis.** Estimate reachable occupancy by stratum, then require every classifier error to reduce to a named missing constructor, environment dependency or timeout class. **Decisive positive result.** A held-out reachability sieve predicts all outcomes and emits checkable witnesses. **Useful null result.** Reachability depends materially on undeclared environment state or resource limits. **Stop rule.** Stop promotion if two pinned rebuilds disagree on any held-out declaration. --- ### EX-002 — Normalization-orbit stability challenge **Frontier:** FA-02 — Representation and normalization stability (quotient) **Mode/readiness:** computational; ready_now **Indicative scale:** 4–8 weeks; proof engineer and data engineer **Topics:** OT-033 **Detector bundles:** DET-001, DET-002 **Platform.** The same declaration corpus expanded into equivalence families by alpha, beta, zeta, metadata, notation and definitional rewrites. **Hypothesis.** A canonical address is constant on declared equivalence orbits while retaining distinctions needed by later detectors. **Interventions** - Generate 20–100 semantics-preserving variants per declaration - Change normalization policy one rule at a time - Insert non-equivalent near neighbours with matched surface statistics **Recorded observables** - Raw and normalized terms - Address mobility matrix - Collision and split rates - Proof of equivalence or non-equivalence **Controls** - Identity transformation - Known beta/zeta pairs - Adversarial near-equivalent terms - Cross-version rebuild **Analysis.** Treat normalization as a quotient experiment: measure false splits within orbits and false merges between orbits with bootstrap intervals by declaration family. **Decisive positive result.** A preregistered policy achieves zero false merges and bounded, explained false splits on held-out families. **Useful null result.** No single policy is stable across source-native families. **Stop rule.** Retain a multi-chart atlas if global canonicalization requires erasing a detector-relevant distinction. --- ### EX-003 — Blind detector-fibre identification challenge **Frontier:** FA-03 — Detector resolution and semantic fibres (detector fold) **Mode/readiness:** computational_hybrid; ready_now **Indicative scale:** 6–10 weeks; modelling, statistics and domain-review trio **Topics:** OT-001, OT-010, OT-019, OT-021, OT-033, OT-034, OT-035, OT-043, OT-044, OT-048, OT-052, OT-054, OT-055 **Detector bundles:** DET-002, DET-003, DET-026, DET-029, DET-030, DET-036, DET-039 **Platform.** A benchmark mixing real declaration trees, driven-oscillator records, five-coordinate incidence records and synthetic mechanism pairs with controlled collisions. **Hypothesis.** Adaptive detector enrichment separates mechanisms at finite cost, or exposes an exact residual equivalence class. **Interventions** - Reveal label, role, rooted depth, phase, sideband and intervention channels sequentially - Hold out complete mechanism families rather than random rows - Adversarially search for record-equivalent parameter pairs **Recorded observables** - Separation depth - Efficient Fisher singular values - Global equivalence-class diameter - Cost per acquired channel **Controls** - Known tangent-degenerate pairs - AZ-057 three-mechanism closed benchmark - Label permutations - Noise and calibration folds **Analysis.** Plot information gained per marginal cost and certify global injectivity only after interval or adversarial optimization over nuisance parameters. **Decisive positive result.** Every declared pair is separated on held-out interventions or assigned to an explicit quotient class. **Useful null result.** A nontrivial model pair remains exactly record-equivalent. **Stop rule.** Do not add more detectors when the residual equality is algebraic rather than noise-limited. --- ### EX-004 — Two-port positive-response cone tomography **Frontier:** FA-04 — Positivity, rank and response cones (cone and rank surface) **Mode/readiness:** hybrid; ready_after_design **Indicative scale:** 3–6 months on an existing microwave or transport platform **Topics:** OT-008, OT-047, OT-054, OT-055 **Detector bundles:** DET-004, DET-019, DET-039 **Platform.** A passive two-port microwave network first, followed by a common-sample particle/energy response implementation. **Hypothesis.** A source-native 2×2 response remains positive and genuinely rank two over a controlled frequency window. **Interventions** - Drive each port independently and coherently - Sweep frequency, temperature and relative phase - Change the declared source basis without refitting physics **Recorded observables** - Complex 2×2 transfer or Kubo matrix - Eigenvalues and determinant - Cross-spectrum phase - Calibration covariance **Controls** - Passive standards - Rank-one synthetic channel - Source-basis congruence - Energy-gauge repartition **Analysis.** Use covariance-aware cone projection; test whether the smaller eigenvalue remains positive after de-embedding and order-of-limits checks. **Decisive positive result.** The second response direction survives calibration, basis and low-frequency scaling. **Useful null result.** Every controlled window becomes rank one or scheme-sheared. **Stop rule.** Stop physical interpretation if passivity violations exceed the calibrated systematic envelope. --- ### EX-005 — Reversal-calibrated orientation lift **Frontier:** FA-05 — Orientation and phase lifts (branched cover) **Mode/readiness:** laboratory; ready_after_design **Indicative scale:** 6–12 months using an integrated-photonic or microwave platform **Topics:** OT-003, OT-039 **Detector bundles:** DET-004, DET-005, DET-007, DET-012, DET-015, DET-017, DET-037 **Platform.** A three-mode interferometric loop with an independently calibrated birefringent, chiral or boundary reference axis. **Hypothesis.** An oriented phase changes sign under physical reversal while even overlap and attenuation do not. **Interventions** - Run forward and reversed loops - Flip the external orientation reference - Remove the reference element while preserving the loop **Recorded observables** - Complex output amplitudes - Wilson-loop or process matrix - Even intensity overlap - Odd phase contrast **Controls** - Reciprocal phase shifter - Matched insertion loss - Basis relabelling - Reference-off device **Analysis.** Fit even and odd channels jointly; the claimed lift is the interaction term between loop reversal and reference reversal, not a fitted absolute phase. **Decisive positive result.** The odd contrast reverses twice as predicted and is stable to gauge choice. **Useful null result.** The signal is removable by basis phase or reproduced by differential loss. **Stop rule.** Reject the orientation claim if the reference cannot be calibrated without selected-state data. **Primary precedents.** [https://doi.org/10.1038/s41467-025-58794-3](https://doi.org/10.1038/s41467-025-58794-3); [https://doi.org/10.1038/s41566-022-00976-2](https://doi.org/10.1038/s41566-022-00976-2) --- ### EX-006 — Three-mode projector-return and coherent-holonomy chip **Frontier:** FA-06 — Projector holonomy and coherent transport (loop and lift) **Mode/readiness:** laboratory; ready_after_design **Indicative scale:** 9–18 months including foundry cycle **Topics:** OT-002, OT-009, OT-010, OT-012, OT-013, OT-037, OT-043, OT-044 **Detector bundles:** DET-004, DET-005, DET-006, DET-013, DET-014 **Platform.** Silicon-nitride or femtosecond-written waveguides implementing a dark three-mode subspace with programmable coupling path. **Hypothesis.** The spectral projector returns after a closed control loop while the coherent frame acquires a nontrivial, order-sensitive holonomy. **Interventions** - Fabricate forward, reverse and path-order-swapped coupling profiles - Inject a tomographically complete input set - Sweep wavelength and deliberately add calibrated splitting and loss **Recorded observables** - Full complex 3×3 transfer matrix - Projector distance - Process singular values - Leakage and insertion loss **Controls** - Zero-area loop - Dynamical-phase echo - Static splitter - Differential-loss mimic **Analysis.** Blindly compare geometric, static-splitting and loss models using held-out wavelengths and reversal data. **Decisive positive result.** Projector closure and reversed noncommuting transport survive independent loss calibration. **Useful null result.** A static or non-Hermitian nuisance model predicts every record. **Stop rule.** Do not fabricate until full-wave tolerancing leaves a resolvable process-matrix margin. **Primary precedents.** [https://doi.org/10.1038/s41467-025-58794-3](https://doi.org/10.1038/s41467-025-58794-3); [https://doi.org/10.1038/s41467-025-62481-8](https://doi.org/10.1038/s41467-025-62481-8) --- ### EX-007 — Selector-on/off basin tomography **Frontier:** FA-07 — Dynamical selection and attraction (basin boundary) **Mode/readiness:** laboratory; ready_after_design **Indicative scale:** 9–18 months; can share EX-006 hardware **Topics:** OT-003, OT-009, OT-013, OT-022, OT-037, OT-039, OT-054 **Detector bundles:** DET-004, DET-006, DET-007, DET-008, DET-011, DET-017, DET-025, DET-039 **Platform.** The holonomy chip or modular acoustic loop extended by a separately addressable dissipative or anisotropic selector segment Y. **Hypothesis.** A carrier-native Y changes the fixed point and contraction tangent rather than merely relabelling a preselected outcome. **Interventions** - Prepare at least twelve initial states spanning the accessible manifold - Toggle Y on/off and reverse its axis - Match total loss and cycle time across conditions **Recorded observables** - Cycle-by-cycle state tomography - Return-map Jacobian - Fixed-point location - Escape and convergence times **Controls** - Loss-matched dummy segment - Y inferred only from independent calibration - Multiple initial states - Axis reversal **Analysis.** Fit a common generative return map with Y as a preregistered intervention and compare contraction spectra, not endpoint images alone. **Decisive positive result.** Y produces the predicted fixed-point shift and subunit tangent eigenvalue across initial states. **Useful null result.** The same contraction occurs with Y absent or after loss matching. **Stop rule.** Stop if Y can only be defined retrospectively from the selected state. --- ### EX-008 — Complete-outcome instrument tomography **Frontier:** FA-08 — Measurement backaction and conditioning (measurement quotient) **Mode/readiness:** laboratory; ready_now_on_existing_hardware **Indicative scale:** 6–12 months on a superconducting qutrit or continuously monitored qubit **Topics:** OT-004, OT-043, OT-052, OT-054, OT-055 **Detector bundles:** DET-008, DET-009, DET-039 **Platform.** A three-level superconducting circuit with single-shot IQ readout, real-time feedback and calibrated missed-record efficiency. **Hypothesis.** Apparent attraction can be separated into unconditional channel dynamics and record-conditioned trajectory branches. **Interventions** - Vary measurement strength and efficiency - Run feedback on/off - Retain every outcome including null and discarded records **Recorded observables** - Raw time-tagged IQ stream - Conditioned trajectories - Unconditional density matrix - Reconstructed quantum instrument **Controls** - Detector tomography - Feedback-disabled channel - Artificial record erasure - State-preparation cross-validation **Analysis.** Reconstruct completely positive outcome maps, sum them to the unconditional channel and propagate efficiency uncertainty. **Decisive positive result.** Conditioned and unconditional maps are separately identifiable and predict held-out sequences. **Useful null result.** Only postselected branches contract while the unconditional manifold remains neutral. **Stop rule.** No selection claim is admitted if unrecorded outcomes can span the effect. **Primary precedents.** [https://doi.org/10.1038/s41586-019-1287-z](https://doi.org/10.1038/s41586-019-1287-z); [https://doi.org/10.1038/s42005-024-01790-8](https://doi.org/10.1038/s42005-024-01790-8) --- ### EX-009 — Matched-contraction stochastic energetics trial **Frontier:** FA-09 — Thermodynamic cost and resource budgets (resource tradeoff) **Mode/readiness:** laboratory; ready_now **Indicative scale:** 4–8 months on an optical-feedback trap **Topics:** OT-005, OT-036, OT-054, OT-055 **Detector bundles:** DET-009, DET-010, DET-039 **Platform.** A 1–2 μm colloidal bead in a programmable double-well optical or feedback potential. **Hypothesis.** Protocols with matched endpoint contraction can have measurably different heat, work, activity and information-reset costs. **Interventions** - Implement equilibrium relaxation, hidden-cycle drive and measurement-feedback protocols - Match endpoint error and completion time - Sweep protocol duration **Recorded observables** - Full bead trajectory - Calibrated potential - Stochastic work and heat - Jump activity and record-reset bits **Controls** - Quasistatic Landauer erasure - No-feedback cycle - Time-reversed protocol - Trap-stiffness calibration **Analysis.** Infer trajectory-level energetics before comparing equal-precision/equal-speed frontiers; keep controller energy as a separate ledger. **Decisive positive result.** Distinct protocols occupy separated precision-speed-cost surfaces despite equal contraction. **Useful null result.** Differences vanish after complete apparatus accounting. **Stop rule.** Do not infer cost from convergence rate alone. **Primary precedents.** [https://doi.org/10.1038/nature10872](https://doi.org/10.1038/nature10872); [https://doi.org/10.1038/s41467-018-03686-y](https://doi.org/10.1038/s41467-018-03686-y) --- ### EX-010 — Rare-event and hysteresis trajectory campaign **Frontier:** FA-10 — Metastability, rare events and hysteresis (large-deviation boundary) **Mode/readiness:** laboratory; ready_now **Indicative scale:** 6–12 months; long unattended acquisition required **Topics:** OT-010, OT-020, OT-023, OT-043, OT-053 **Detector bundles:** DET-007, DET-008, DET-011, DET-013, DET-037 **Platform.** An injection-locked VCSEL, Kerr resonator or nanomagnetic element with tunable bistability and single-event readout. **Hypothesis.** Waiting-time tails, sweep hysteresis and reversal asymmetry distinguish metastable selection from a unique deterministic attractor. **Interventions** - Sweep control in both directions at logarithmically spaced rates - Hold inside the bistable region for long dwell windows - Apply calibrated weak perturbation pulses **Recorded observables** - Every switching time - Full drive and output trace - Residence-time distributions - Ringdown and sidebands **Controls** - Noise-floor injection - Monostable operating points - Randomized sweep order - Detector-dead-time audit **Analysis.** Fit survival curves and large-deviation rate functions jointly; reserve late trajectories for posterior predictive tests. **Decisive positive result.** One mechanistic landscape predicts rates, hysteresis width and perturbation response across sweeps. **Useful null result.** Observed memory is detector lag or ordinary bandwidth filtering. **Stop rule.** Stop adding dwell time once confidence intervals are dominated by nonstationary drift rather than rare-event count. --- ### EX-011 — Tunable finite-chain pole-to-cut spectrometer **Frontier:** FA-11 — Spectral boundary dynamics (pole, cut and threshold) **Mode/readiness:** laboratory; ready_now **Indicative scale:** 4–9 months with microwave or acoustic resonators **Topics:** OT-012, OT-013, OT-014, OT-036, OT-041, OT-051, OT-052, OT-053 **Detector bundles:** DET-006, DET-013, DET-014, DET-022, DET-023, DET-037 **Platform.** A 10–40 site coupled-resonator chain with one tunable side-coupled defect and access to both end-port scattering and local density of states. **Hypothesis.** Bound states, edge collisions, resonances, phase winding and time delay form one branch-consistent ledger on the same carrier. **Interventions** - Tune defect frequency and coupling through both certified pole collisions - Repeat for several chain lengths and losses - Excite locally and from both scattering ports **Recorded observables** - Complex S-matrix - Local spectra - Ringdown - Pole fits and phase winding **Controls** - Disconnected defect - Uniform chain - Known rank-one impurity - Synthetic branch-choice challenge **Analysis.** Fit a causal rational model globally, continue poles with uncertainty and compare the finite-band Levinson ledger before extrapolating length. **Decisive positive result.** Every spectral stratum and both collision laws agree without changing the model family. **Useful null result.** Crossings depend on sampling, fit order or Riemann-sheet convention. **Stop rule.** Freeze the result at the finite carrier if loss prevents stable continuation. --- ### EX-012 — Same-film tensor-to-Casimir closure **Frontier:** FA-12 — Material response and fluctuation forces (response surface) **Mode/readiness:** laboratory; multi_stage **Indicative scale:** 18–36 months across materials, transport, optics and force metrology teams **Topics:** OT-006, OT-007, OT-011 **Detector bundles:** DET-012, DET-013, DET-014, DET-015, DET-016, DET-017, DET-018, DET-029 **Platform.** One batch of anisotropic magnetic films split into witness coupons and a force/torque device, with common growth and metrology records. **Hypothesis.** Broadband causal response inferred without force data predicts a differential Casimir force or torque after all ordinary nuisances are propagated. **Interventions** - Measure Hall, Seebeck, Nernst and magneto-optical response versus temperature, field and angle - Freeze a causal tensor posterior - Measure angle- and distance-dependent parallel-minus-antiparallel force/torque **Recorded observables** - Transport and ellipsometry tensors - Film thickness, roughness and domains - Force/torque curves - Patch and magnetic backgrounds **Controls** - Nonmagnetic sibling film - Orientation reversal - Multiple separations - Blind force analysis and dummy spacing scans **Analysis.** Propagate the pre-force tensor posterior through Lifshitz theory with a full nuisance covariance; prohibit response refitting to force data. **Decisive positive result.** The preregistered differential curve is recovered with nuisance alternatives disfavoured. **Useful null result.** Optical-tail, patch, roughness or magnetic freedom spans the effect. **Stop rule.** Do not proceed to force fabrication until OT-006 cross-channel prediction passes on held-out films. **Primary precedents.** [https://doi.org/10.1103/RevModPhys.81.1827](https://doi.org/10.1103/RevModPhys.81.1827); [https://doi.org/10.1103/PhysRevLett.119.183001](https://doi.org/10.1103/PhysRevLett.119.183001) --- ### EX-013 — Common-sample multichannel Onsager-rank experiment **Frontier:** FA-13 — Multichannel transport and hydrodynamic rank (rank surface and order of limits) **Mode/readiness:** laboratory_computational; ready_after_design **Indicative scale:** 12–24 months including microfabrication **Topics:** OT-008, OT-016, OT-040, OT-044, OT-047 **Detector bundles:** DET-012, DET-013, DET-019, DET-020, DET-021, DET-029 **Platform.** A suspended four-terminal microbridge allowing independent electrical and thermal biases, magnetic-field reversal and AC thermometry. **Hypothesis.** The particle/energy response has a stable second eigen-direction in a declared low-frequency transport window. **Interventions** - Independently modulate voltage and heater power - Sweep frequency, temperature, field, filling and contact geometry - Reverse injection/emission contacts **Recorded observables** - Complex charge and heat currents - Full 2×2 Onsager matrix - Noise cross spectra - Thermal and electrical calibration covariance **Controls** - Onsager-Casimir reciprocity - Open/short thermal standards - Energy-zero shift - Contact-swap reversal **Analysis.** Infer eigenvalues with errors in both currents and forces; test limit ordering and compare with common-sample KPM/typicality. **Decisive positive result.** Rank two and reciprocity survive frequency-window, size and energy-gauge changes. **Useful null result.** The second direction is contact leakage or exits every controlled low-frequency window. **Stop rule.** Stop rank claims when thermal cross-talk exceeds the smaller singular value. **Primary precedents.** [https://arxiv.org/abs/1306.3694](https://arxiv.org/abs/1306.3694) --- ### EX-014 — Defect information beyond mode count **Frontier:** FA-14 — Topology, defects and protected transport (topological wall) **Mode/readiness:** laboratory; ready_now **Indicative scale:** 6–12 months with acoustic or topoelectric resonator arrays **Topics:** OT-015, OT-046, OT-051 **Detector bundles:** DET-023, DET-024 **Platform.** A reconfigurable 2D acoustic or circuit lattice containing matched trivial, domain-wall, partial-dislocation and threefold-quotient defects. **Hypothesis.** Defects with the same protected mode count remain distinguishable by oriented transport, spatial texture and scattering phase. **Interventions** - Move source and detector around the defect - Reverse helicity or circulation - Add controlled symmetry-preserving and symmetry-breaking disorder **Recorded observables** - Site-resolved complex field - Local density of states - Scattering matrix - Winding, delay and localization length **Controls** - Defect-free lattice - Same-index inequivalent defects - Boundary-condition variants - Blind geometric relabelling **Analysis.** Condition on equal index and compare the remaining response tensor; demand a held-out observable beyond mode count. **Decisive positive result.** A non-index record robustly tracks defect geometry and quotient operation. **Useful null result.** All extra information is arbitrary boundary data. **Stop rule.** Retain only the topological count if disorder destroys the extra record before the bulk gap closes. **Primary precedents.** [https://doi.org/10.1038/s41467-022-29785-5](https://doi.org/10.1038/s41467-022-29785-5); [https://doi.org/10.1038/s41467-024-50666-6](https://doi.org/10.1038/s41467-024-50666-6) --- ### EX-015 — Disorder–spread–coarse-graining scaling tube **Frontier:** FA-15 — Disorder, causal spread and scale (scaling frontier) **Mode/readiness:** laboratory_computational; ready_after_design **Indicative scale:** 6–12 months on a programmable resonator or qubit array **Topics:** OT-011, OT-019, OT-021, OT-022, OT-023, OT-051, OT-054, OT-055 **Detector bundles:** DET-011, DET-012, DET-014, DET-015, DET-020, DET-021, DET-024, DET-025, DET-026, DET-030, DET-039 **Platform.** Arrays of 16–128 microwave/acoustic resonators or available quantum-simulator sites with programmable onsite disorder and couplings. **Hypothesis.** A small set of propagation and memory invariants remains stable across disorder, size, time and coarse-graining. **Interventions** - Quench a local excitation - Sweep disorder distribution and correlation length - Repeat across array sizes and detector aggregation scales **Recorded observables** - Space-time response cone - Out-of-cone leakage - Participation ratio - Memory and return statistics **Controls** - Clean ballistic array - Strongly localized array - Detector convolution - Randomized disorder seeds **Analysis.** Fit a joint finite-size/time scaling model and test semigroup consistency between raw and coarse observations. **Decisive positive result.** One invariant scaling tube predicts held-out size and disorder ensembles. **Useful null result.** Every apparent bound is detector- or size-specific. **Stop rule.** Do not extrapolate if fitted constants drift monotonically with the largest available size. --- ### EX-016 — Code-island operator-law falsification **Frontier:** FA-16 — Arithmetic and code islands (arithmetic cliff) **Mode/readiness:** exact_computational; ready_now **Indicative scale:** 2–6 months; algebra, coding and proof expertise **Topics:** OT-017, OT-025, OT-045, OT-050 **Detector bundles:** DET-027, DET-028 **Platform.** The finite Monster-address, Golay/Leech-neighbour and 24-dimensional form data already represented in the corpus. **Hypothesis.** At least one source-derived operator family—not labels alone—satisfies preregistered Hecke, modular-trace or product-preserving relations. **Interventions** - Define operators before inspecting occupancy coincidences - Apply them to the full finite atlas - Compare integral, rational and mod-p reductions **Recorded observables** - Exact operator matrices - Relation residuals - Graded traces - Integrality and orbit certificates **Controls** - Random label permutations - Same divisor multiset with destroyed multiplication - Known Hecke toy module - Independent exact implementation **Analysis.** Require exact identities over integers or declared rings; numerical near-equality has no evidential weight. **Decisive positive result.** A nontrivial operator law holds without fitting prime-address occupancy. **Useful null result.** Only numerology, dimensions or divisor labels survive. **Stop rule.** Close the branch after all natural source-derived multiplications fail the preregistered identities. --- ### EX-017 — Two-carrier commuting-square hardware-in-the-loop test **Frontier:** FA-17 — Operator and carrier translation (translation gap) **Mode/readiness:** hybrid; staged **Indicative scale:** 12–24 months after a source-native map is written **Topics:** OT-017, OT-024, OT-025, OT-026, OT-027, OT-035, OT-038, OT-039, OT-042, OT-045, OT-046, OT-047, OT-048, OT-049, OT-050, OT-051, OT-053, OT-054 **Detector bundles:** DET-027, DET-035, DET-036, DET-038, DET-039 **Platform.** A photonic three-mode chip and an acoustic or microwave three-port emulator implementing the same declared A2 operation in native units. **Hypothesis.** A provenance-bearing transformation transports projector, complex structure, losses and observable records between carriers while preserving composition. **Interventions** - Prepare matched labelled states in both carriers - Apply generators singly and in compositions - Transport predictions in both directions without basis refitting **Recorded observables** - Native process matrices - Unit and label dictionary - Composition residual - Irreversible loss ledger **Controls** - Unlabelled matrix similarity - Fitted basis alignment - One-generator-only agreement - Held-out composite paths **Analysis.** Test path equality in the commuting square on held-out states, with separate structural and calibrated physical error budgets. **Decisive positive result.** Both paths agree and the map predicts a native observable in the second carrier. **Useful null result.** Agreement exists only after erasing labels, units or orientation. **Stop rule.** Do not build the second device until the source-side map preserves products symbolically. --- ### EX-018 — Uniform-limit adversarial scattering benchmark **Frontier:** FA-18 — Consistency, scattering and continuum control (consistency boundary) **Mode/readiness:** theorem_computation; ready_now **Indicative scale:** 6–18 months depending on whether a theorem survives **Topics:** OT-014, OT-016, OT-018, OT-021, OT-031, OT-040, OT-046, OT-051, OT-052, OT-053, OT-054, OT-055 **Detector bundles:** DET-016, DET-022, DET-023, DET-030, DET-034, DET-039 **Platform.** Families of finite chains, lattice gauge Hamiltonians and D4 defect operators with controllable volume, regulator, time and observable class. **Hypothesis.** A proposed positivity, gap or admissibility bound has constants uniform in every declared limiting direction. **Interventions** - Search adversarial sequences jointly over volume, regulator and boundary condition - Enlarge the observable class - Compare pole, phase, Euclidean and real-time estimators **Recorded observables** - Best lower bound by scale - Extremizing sequence - Finite-size drift - Proof-assistant-checkable inequalities where available **Controls** - Known gapped and gapless models - Order-of-limits swaps - Boundary-localized false gaps - Synthetic positivity violations **Analysis.** Attempt to falsify uniformity first; a successful result must supply explicit constants and domain assumptions. **Decisive positive result.** One bound remains positive and stable under all declared limits. **Useful null result.** An admissible sequence drives the bound to zero or changes its sign. **Stop rule.** Record a scoped no-go once a countersequence is certified; do not replace it with larger finite numerics. --- ### EX-019 — Joined torsion–flavour–cosmology likelihood with a polarized null experiment **Frontier:** FA-19 — Flavour, gravity and cosmology (fundamental firewall) **Mode/readiness:** conditional_hybrid; conditional_on_action **Indicative scale:** 12–36 months after a frozen action and matching packet exist **Topics:** OT-028, OT-029, OT-030, OT-031, OT-042, OT-047, OT-048, OT-049, OT-050, OT-052 **Detector bundles:** DET-029, DET-031, DET-032, DET-033, DET-034, DET-035, DET-036 **Platform.** A rotating unpolarized source mass near a K–Rb–Ne comagnetometer or polarized spin sensor, joined to collider/flavour and cosmological public likelihoods through one action. **Hypothesis.** One protected parameter map predicts range, species ratios and signs across laboratory, flavour and cosmological observables. **Interventions** - Modulate source distance, velocity and orientation - Reverse sensor polarization and source rotation - Freeze threshold matching before opening laboratory data **Recorded observables** - Time-series pseudomagnetic field - Environmental witness channels - Flavour/collider likelihood terms - Expansion, lensing and structure likelihoods **Controls** - Nonmagnetic dummy source - Blind rotation phase - Multiple source distances - Generic pseudoscalar comparison model **Analysis.** Perform one hierarchical likelihood with shared parameters and explicit theory errors; report the generic-EFT fit beside the protected model. **Decisive positive result.** A common no-refit parameter point survives all channels and predicts a held-out reversal or range curve. **Useful null result.** Generic free pseudoscalar coefficients fit equally well or one dataset requires an incompatible parameter value. **Stop rule.** No apparatus-specific interpretation begins until the action fixes the measured operator and species ratios. **Primary precedents.** [https://arxiv.org/abs/2203.07050](https://arxiv.org/abs/2203.07050); [https://arxiv.org/abs/2408.15691](https://arxiv.org/abs/2408.15691) --- ### EX-020 — Exceptional-model end-to-end red-team **Frontier:** FA-20 — Exceptional and unification interpretation (fundamental firewall) **Mode/readiness:** theory_computation; conditional_on_model **Indicative scale:** 12–36 months for one fully specified candidate **Topics:** OT-026, OT-032, OT-038, OT-042, OT-045, OT-049, OT-050 **Detector bundles:** DET-028, DET-035 **Platform.** A symbolic and numerical pipeline accepting one exceptional-group action, representation dictionary, vacuum and symmetry-breaking chain. **Hypothesis.** A single candidate can pass anomaly, Hessian, gauge-spectrum, threshold and observable gates without freely chosen embeddings. **Interventions** - Enumerate the full chiral spectrum and anomaly polynomial - Compute the complete scalar Hessian and gauge masses including coloured directions - Run RG and threshold matching to a frozen low-energy likelihood **Recorded observables** - Anomaly coefficients - Hessian eigenpairs - Gauge and matter spectrum - Running couplings and held-out observable predictions **Controls** - Basis and embedding changes - Known inconsistent truncations - Parameter-count audit - Generic low-energy EFT baseline **Analysis.** Use a sequential gate: failure at any exact consistency stage prevents phenomenological fitting; successful candidates are compared by predictive compression. **Decisive positive result.** One anomaly-free stable vacuum produces a no-refit low-energy prediction that beats the independent EFT. **Useful null result.** The construction relies on dimension matches, arbitrary embeddings or compensating free parameters. **Stop rule.** Terminate a candidate immediately on anomaly, ghost, unstable coloured direction or underdetermined matching. --- ### EX-021 — Source-complete W-associated amplitude and meson validation **Frontier:** FA-19 — Flavour, gravity and cosmology (sixth-coordinate detector quotient) **Mode/readiness:** theory_kernel_then_public_data_recast; source_complete_amplitude_required **Indicative scale:** 2–6 months for the amplitude audit; 3–9 months for a public-likelihood recast after release **Topics:** OT-047, OT-048 **Detector bundles:** DET-029, DET-036 **Platform.** A gauge-complete W-associated ALP amplitude with independently signed current and direct-boson sources, followed by Belle/Belle II or successor B-to-K(*)a validation in the same EFT basis. **Hypothesis.** A differential W-associated interference shape survives source completion and is not proportional to the weak quotient (-1/2,1), while the meson arm consistently constrains the ctilde_WW/current/gluon/UV-flavor combination identified by AZ-081. **Interventions** - Run AZ-074's 27 signed source settings on common phase-space points - Ablate direct-boson, top-current and bottom-current diagrams only in gauge-complete subsets - Scan the ALP mass and lifetime on one shared grid - Separate charged/neutral B and K/K* categories - Hold out at least one decay or collider category until the joint coefficients are frozen **Recorded observables** - Differential W-associated quadratic response coefficients and Ward residuals - Source-complete (A_2,c_WW) singular spectrum by phase-space bin - B-to-K(*) diphoton, dilepton or missing-energy binned likelihood - Cross-scale held-out category prediction **Controls** - Published Eq. (74) rank-one MFV kernel - Isolated c_WW matching-term false-positive comparator - One-coupling g_aW interpretation - Current-only c_WW=0 hypothesis - Mass sidebands and lifetime/acceptance closure - CKM and form-factor covariance **Analysis.** Complete every coefficient through one frozen basis, running and weak-threshold dictionary before calculating rank. Test Ward identities and the 27-setting quadratic closure at differential level; only a non-collinear surviving shape proceeds to detector folding. Use the meson likelihood as a common-EFT validation arm and profile current, gluonic, CKM, hadronic, decay and detector nuisances. **Decisive positive result.** A source-complete non-collinear differential shape clears its preregistered singular-value threshold, survives detector folding and predicts the held-out meson or collider category without refitting. **Useful null result.** The full amplitude recombines to the weak quotient, as the MFV meson kernel does, closing the proposed sixth-coordinate route while retaining a useful cross-scale constraint. **Stop rule.** Do not claim separation from an isolated matching term, one-coupling limit or diagram subset; stop unless source completion, Ward checks, held-out prediction and the full covariance all preserve the second shape. **Primary precedents.** [https://arxiv.org/abs/2012.12272](https://arxiv.org/abs/2012.12272); [https://arxiv.org/abs/2507.01249](https://arxiv.org/abs/2507.01249) --- ### EX-022 — Single-ion operator-opening and lifetime ledger **Frontier:** FA-11 — Spectral boundary dynamics (operator-to-record admissibility boundary) **Mode/readiness:** laboratory_atomic_spectroscopy; ready_on_established_platform **Indicative scale:** 6–12 months on an equipped trapped-ion platform; helium transfer arm is a separate specialist campaign **Topics:** OT-036 **Detector bundles:** DET-013, DET-014, DET-037 **Platform.** A single trapped calcium-40 ion prepared in the metastable 3d 2D5/2 level through the 729 nm electric-quadrupole line, with an independently calibrated 854 nm electric-dipole quench and state-dependent fluorescence readout. **Hypothesis.** One typed operator model predicts the added decay rate, detuning curve, angular dependence and helicity reversal when a calibrated electric-dipole disposal channel is opened from an otherwise metastable level. **Interventions** - Sweep 854 nm intensity and detuning in the weak unsaturated regime - Rotate the polarization relative to the magnetic quantization axis - Reverse helicity and resolve addressed Zeeman sublevels - Vary observation time while holding state preparation and fluorescence thresholds fixed - Repeat intrinsic-lifetime measurements with the quench beam physically blocked **Recorded observables** - State-survival curves and inferred partial and total rates - Polarization- and sublevel-resolved photon counts - Frequency-domain line shape and time-domain lifetime - Branching fractions, quench power, detuning and magnetic-field metadata - Background-gas, leakage-light, motional, magnetic-noise and detector-dead-time witnesses **Controls** - Blocked-quench intrinsic-lifetime control - Off-resonant and wrong-polarization controls - Known Clebsch–Gordan angular ratios - Collision and stray-light pressure/power sweeps - Independent power, polarization, frequency and fluorescence calibration **Analysis.** Fit all survival, detuning, orientation and helicity records with one level Hamiltonian, one reduced matrix element and shared detector/nuisance parameters. In the weak regime require the added rate to begin linearly in intensity with the predicted tensor-component ratios; do not infer a microscopic lifetime from linewidth alone. **Decisive positive result.** The same operator model predicts the added rate, reversal, angular and spectral records on held-out settings while independently bounded collision, leakage and detector models remain below the effect. **Useful null result.** No added rate is resolved at the calibrated sensitivity, or a conventional collision, leakage, saturation, dephasing or detector model spans the record without the proposed operator opening. **Stop rule.** Stop operator interpretation if the intrinsic lifetime is not reproduced, the quench field is not independently calibrated, time- and frequency-domain widths disagree outside the declared dephasing model, or polarization leakage spans the tensor signature. **Primary precedents.** [https://doi.org/10.1103/PhysRevA.91.022511](https://doi.org/10.1103/PhysRevA.91.022511); [https://doi.org/10.1103/RevModPhys.84.175](https://doi.org/10.1103/RevModPhys.84.175); [https://doi.org/10.1103/PhysRevLett.125.013002](https://doi.org/10.1103/PhysRevLett.125.013002) **Reference bridge.** [AZ-127](../admissibility/admissibility_127_transition_operators_selection_rules_and_lifetime_frontier.md) ## Execution order 1. **Immediate, no new apparatus:** EX-001, EX-002, EX-003, EX-016 and EX-018. 2. **Existing-platform pilots:** EX-004, EX-008, EX-009, EX-010, EX-011, EX-014 and EX-022. 3. **Shared-device build programme:** EX-005, EX-006 and EX-007, followed by EX-017. 4. **Materials and transport campaign:** EX-012, EX-013 and EX-015. 5. **Model-gated fundamental tests:** EX-019 and EX-020 only after their action/dictionary entry conditions are met. The shared-device sequence is deliberate: EX-005–EX-007 can use one three-mode carrier, while EX-012 forces the response tensor to be frozen before the Casimir record is opened. These dependencies make null results informative and limit refitting freedom.