Runnable scientific previews
FusionSim, Nuclear Data, Mirror Reactor, Torus Orbit, AlphaTrack, and DirectConvert retain distinct intended uses and claim boundaries.
Research Preview 1.0 · 21 August 2026
Small models. Explicit evidence. Bounded university asks.
HBF Network publishes reduced, inspectable p–¹¹B research and education software. This release consolidates the public portfolio into six scientific modules, two lessons, one evidence system, and three collaboration work packages.
Claim boundary: HBF does not operate a reactor. These tools are not reactor designs, net-energy claims, experimental results, digital twins, or engineering predictions. Internal checks do not equal independent scientific review or experimental validation.
First partnership release
Two runnable contributions; one ask at a time
The primary collaboration path is an independent reproduction of one p–¹¹B reactivity model. A detector laboratory may instead co-design the AlphaTrack real-CR-39 validation pilot. The other modules remain supporting demonstrations.
Public maturity statement
Internal software and numerical checks passed for the declared cases. No independent scientific or experimental validation.
See each evidence dimension separately, including source pedigree, numerical verification, literature reproduction, review, and experimental-validation state.
Portfolio architecture
One navigable scientific system
Concepts before controls
A normalized Gamow-shape lesson and a fusion-configuration primer establish vocabulary without implying device performance.
Status is multidimensional
No maturity badge collapses software tests, numerical convergence, literature agreement, review, experiments, and data pedigree.
Research integrity
Designed to preserve negative results and unresolved differences
Reproducible identity
Versioned RunSpec/ResultSpec schemas, canonical JSON and SHA-256 identifiers, source/model records, fixtures, tolerances, release notes, and machine-readable evidence states ship with the package.
Validation-domain discipline
Verification asks whether software and mathematics meet declared requirements. Validation asks how well a model represents a real-world referent for a stated use and domain. This release does not blur the two.