Partnership principles
- No reactor-performance or net-energy claim.
- No outcome is predetermined.
- The partner chooses or approves the independent method.
- Negative results and abstentions are retained and published.
- Conflicts, funding and roles are disclosed.
- Research outputs receive durable identifiers and contributor credit.
- Data, IP and publication terms are agreed before any data transfer.
- No token purchase, wallet connection or financial participation is required or mentioned.
Three bounded work packages
WP1 · package ready
Independent p–¹¹B reactivity reproduction
Question. Can an independent implementation reproduce HBF's declared reactivity curve within preregistered tolerances over the supported domain?
- Partner profile
- Computational plasma or nuclear physics group; research software engineer; reaction-data specialist.
- Partner provides
- Independent implementation, source interpretation, numerical comparison, discrepancy classification, signed review.
- HBF provides
- Frozen source list, exact equations and normalisation, source snapshots, run specification, expected values, tolerance rationale, environment, report template.
- Why this one first
- It needs no physical device, produces a publishable record even when the result is negative, and exercises the parts of HBF that are actually strong.
- Stopping rule
- A material source or normalisation ambiguity pauses numerical comparison and becomes a published discrepancy.
- Accountable owner
- Not yet published
Open the WP1 reproduction package
WP2 · package ready
Real CR-39 pilot
Question. Does the AlphaTrack QC and detection workflow remain useful on a preregistered, plate-separated real-data pilot without making unsupported species, energy or yield claims?
- Partner profile
- High-energy-density or laser-plasma diagnostics group with real CR-39 plates, protocol metadata and ground-truth capability.
- Partner provides
- Lawfully shareable images or controlled access, etch and acquisition metadata, calibration context, blinded labels.
- HBF provides
- Synthetic baseline, pipeline, QC and abstention rules, dataset template, blinded evaluation harness, error-analysis interface.
- Governance
- Data-use agreement, background and foreground IP, embargo, access list, retention and deletion, publication review, image licence.
- Stopping rule
- Missing acquisition or calibration metadata, unacceptable leakage, or out-of-domain input requires abstention rather than metric publication.
- Accountable owner
- Not yet published
Review the synthetic baseline
WP3 · package ready
Independent orbit and solver verification
Question. Do HBF trajectories and invariants agree with an independent solver for frozen analytic-field cases within agreed timestep and convergence tolerances?
- Partner profile
- Plasma computation group with an independent orbit integrator, mirror or torus expertise, or an analytic benchmark suite.
- Partner provides
- Independent solver choice, blind outputs, error metrics, review of domain and fidelity statements.
- HBF provides
- Field definitions, units, seeds, integrator settings, reference cases, outputs, convergence tests, run schema.
- Stopping rule
- Coordinate or unit ambiguity, or a non-equivalent field definition, becomes a discrepancy before any trajectory comparison.
- Accountable owner
- Not yet published
Review the orbit module
Honest status of these packages.
All three are prepared: the frozen questions, artifacts, tolerances, stopping rules and decision templates exist and are internally validated. None has been executed. The count of completed independent external decisions is zero, and it will stay zero on every page of this site until a genuine independent party produces one. Named accountable owners and any response-time commitment are also not yet published, because those are organisational facts rather than software.
Institutional intake
Deliberately short and privacy-minimised. Do not attach datasets, images or anything sensitive: no agreement is in place yet, and this form is not a lawful route for research data.