KODEX — KDRIVE (RL-control). reinforcement-learning controller — a feedback policy learned by the cross-entropy method in the twin plant loop; clamped by KGATEBenchmark: RL policy (cross-entropy): tracking **0.0551** vs naive 0.1225, twin-in-the-loopPart of KODEX, the Kronos Family of Codes — a benchmarked AI/ML fusion surrogate suite with calibrated uncertainty and an abstention gate (predict(x) → (y, uncertainty, in_domain)). Provenance: references 10.5281/zenodo.21842371. Home: https://kronosfusionenergy.com/kodex/kdrive · suite: https://github.com/KronosFE/kronos-ml. Honest card preserved verbatim.
KODEX — KDRIVE (RL-control). reinforcement-learning controller — a feedback policy learned by the cross-entropy method in the twin plant loop; clamped by KGATE Benchmark: RL policy (cross-entropy): tracking **0.0551** vs naive 0.1225, twin-in-the-loop Part of KODEX, the Kronos Family of Codes — a benchmarked AI/ML fusion surrogate suite with calibrated uncertainty and an abstention gate (predict(x) → (y, uncertainty, in_domain)). Provenance: references 10.5281/zenodo.21842371. Home: https://kronosfusionenergy.com/kodex/kdrive · suite: https://github.com/KronosFE/kronos-ml. Honest card preserved verbatim.
Priyanca Ford· Zenodo (CERN European Organi...· 0 citations
KODEX — KDRIVE (RL-control). reinforcement-learning controller — a feedback policy learned by the cross-entropy method in the twin plant loop; clamped by KGATE Benchmark: RL policy (cross-entropy): tracking **0.0551** vs naive 0.1225, twin-in-the-loop Part of KODEX, the Kronos Family of Codes — a benchmarked AI/ML fusion surrogate suite with calibrated uncertainty and an abstention gate (predict(x) → (y, uncertainty, in_domain)). Provenance: references 10.5281/zenodo.21842371. Home: https://kronosfusionenergy.com/kodex/kdrive · suite: https://github.com/KronosFE/kronos-ml. Honest card preserved verbatim.
Priyanca Ford· Zenodo (CERN European Organi...· 0 citations
KODEX — KDRIVE (RL-control). reinforcement-learning controller — a feedback policy learned by the cross-entropy method in the twin plant loop; clamped by KGATEBenchmark: RL policy (cross-entropy): tracking **0.0551** vs naive 0.1225, twin-in-the-loopPart of KODEX, the Kronos Family of Codes — a benchmarked AI/ML fusion surrogate suite with calibrated uncertainty and an abstention gate (predict(x) → (y, uncertainty, in_domain)). Provenance: references 10.5281/zenodo.21842371. Home: https://kronosfusionenergy.com/kodex/kdrive · suite: https://github.com/KronosFE/kronos-ml. Honest card preserved verbatim.
KODEX — KGEN (generative). generative surrogate — a PCA/latent model that samples plausible equilibrium flux fields; diffusion is the roadmap upgrade Benchmark: generative (PCA latent, dim 8); samples plausible equilibrium fields (diffusion = roadmap) Part of KODEX, the Kronos Family of Codes — a benchmarked AI/ML fusion surrogate suite with calibrated uncertainty and an abstention gate (predict(x) → (y, uncertainty, in_domain)). Provenance: references 10.5281/zenodo.21842371. Home: https://kronosfusionenergy.com/kodex/kgen · suite: https://github.com/KronosFE/kronos-ml. Honest card preserved verbatim.
Priyanca Ford· Zenodo (CERN European Organi...· 0 citations
In a deuterium–helium-3 (D–^3He) tandem-mirror burner the fusion-born 14.68 MeV proton and 3.67 MeV alpha carry the overwhelming majority of the released power as fast, mirror-trapped ions. Left to thermalise, that energy is shared between electron drag and a broadly heated ion background whose confined, usefully recoverable share is small; the rest exits the loss cone as heat and inflates the recirculating power that sets the burner's engineering gain Q_E. RF alpha-channeling—a resonant wave that extracts perpendicular energy from a fusion product, hands it to the fuel ions, and simultaneously diffuses the spent ash toward the loss boundary—converts this loss into a directed gain. We formalise the mechanism as a bounce-averaged Fokker–Planck problem closed by a quasilinear cyclotron-resonant diffusion operator, derive the Fisch–Rax energy–position coupling E/=/n that makes channeling a bounded lever rather than a closure mechanism, and evaluate it at the frozen M-45 burner operating point (T_i=90 keV, n_e=2.6×10^20 m^-3, x_^3He=0.30, mirror-throat field B_m=17 T, central-cell field B_0c=5.50 T, effective mirror ratio R_mc=4.61). A reduced two-dimensional velocity-space kinetic solve, cross-checked against ray-tracing and particle-in-cell codes on the NVIDIA GPU HPC campaign, raises the recoverable charged fraction from a collisional baseline near 8% to about 80% and recovers up to 195 MW into the confined fuel ions—enough to cover the 40–110 MW plug-localised warm-fill RF cost with margin, while a full-volume warm fill (3.7 GW, 85% of P_ fus) is energetically precluded. We are explicit that this closes only under plug localisation (channeled power confined to 1–3% of the central-cell volume) and that the plant gain remains gated by the plug potential, not by channeling: the design-point Q_E=1.318 is untouched. Every headline quantity is a frozen anchor in the Kronos de-risking register.Key results (frozen anchors): n_e = 2.6 ×10^20; Frec = 8 %; f_n = 5.44 %.Live verification: 11 gate validator(s) with live-recompute cards (9 reproduced, 2 revised). See the Verification section and data/verification.csv.Related: Paper page · De-risking register · 3D model · Learn more about KronosPart of the 2026 Kronos publication series; independently re-run and stamped in the Kronos de-risking register (DOI 10.5281/zenodo.22645689).All numerical values are frozen design-point anchors; see the register.Public research artifact. No proprietary, financial, or supply-chain information is included.Note (REPLACE): This record replaces and supersedes DOI 10.5281/zenodo.22132168; please cite this version.
Priyanca Ford, P I Ford, G L Kulcinski· Figshare· 0 citations
In a deuterium–helium-3 (D–^3He) tandem-mirror burner the fusion-born 14.68 MeV proton and 3.67 MeV alpha carry the overwhelming majority of the released power as fast, mirror-trapped ions. Left to thermalise, that energy is shared between electron drag and a broadly heated ion background whose confined, usefully recoverable share is small; the rest exits the loss cone as heat and inflates the recirculating power that sets the burner's engineering gain Q_E. RF alpha-channeling—a resonant wave that extracts perpendicular energy from a fusion product, hands it to the fuel ions, and simultaneously diffuses the spent ash toward the loss boundary—converts this loss into a directed gain. We formalise the mechanism as a bounce-averaged Fokker–Planck problem closed by a quasilinear cyclotron-resonant diffusion operator, derive the Fisch–Rax energy–position coupling E/=/n that makes channeling a bounded lever rather than a closure mechanism, and evaluate it at the frozen M-45 burner operating point (T_i=90 keV, n_e=2.6×10^20 m^-3, x_^3He=0.30, mirror-throat field B_m=17 T, central-cell field B_0c=5.50 T, effective mirror ratio R_mc=4.61). A reduced two-dimensional velocity-space kinetic solve, cross-checked against ray-tracing and particle-in-cell codes on the NVIDIA GPU HPC campaign, raises the recoverable charged fraction from a collisional baseline near 8% to about 80% and recovers up to 195 MW into the confined fuel ions—enough to cover the 40–110 MW plug-localised warm-fill RF cost with margin, while a full-volume warm fill (3.7 GW, 85% of P_ fus) is energetically precluded. We are explicit that this closes only under plug localisation (channeled power confined to 1–3% of the central-cell volume) and that the plant gain remains gated by the plug potential, not by channeling: the design-point Q_E=1.318 is untouched. Every headline quantity is a frozen anchor in the Kronos de-risking register.Key results (frozen anchors): n_e = 2.6 ×10^20; Frec = 8 %; f_n = 5.44 %.Live verification: 11 gate validator(s) with live-recompute cards (9 reproduced, 2 revised). See the Verification section and data/verification.csv.Related: Paper page · De-risking register · 3D model · Learn more about KronosPart of the 2026 Kronos publication series; independently re-run and stamped in the Kronos de-risking register (DOI 10.5281/zenodo.22645689).All numerical values are frozen design-point anchors; see the register.Public research artifact. No proprietary, financial, or supply-chain information is included.Note (REPLACE): This record replaces and supersedes DOI 10.5281/zenodo.22132168; please cite this version.
Priyanca Ford, P I Ford, G L Kulcinski· Figshare· 0 citations
An autonomous controller can act only on the state it can see, and in a compact spherical tokamak the state must be produced faster than the plant evolves, from a diagnostic set that survives a 14 neutron field and keeps working when individual channels drop out. We formalise the Hyperion breeder diagnostic suite — magnetics, Thomson scattering, interferometry, bolometry, and a neutron camera — and its real-time state estimator as a Bayesian inverse problem, write the forward observation model of each diagnostic and the maximum-a-posteriori estimator with its posterior covariance, and cast observability of the internal state as the rank and conditioning of the associated Fisher information. The estimator's learned core is reduced to practice at the metrics that gate it into service. A graph-neural-network reconstructor, which we identify as an amortised posterior-mean estimator over the sensor graph, recovers the plasma state and a quench precursor jointly at an area under the receiver-operating-characteristic curve (AUC) of 0.980 at a reconstruction normalised RMS error of 0.396, and degrades gracefully rather than catastrophically as channels are removed. A multi-modal fusion of a fibre-Bragg strain-rate channel with a magnetization channel reaches 0.992 (against 0.976 and 0.921 alone) and a true-positive rate of 0.857 at a 1% false-positive rate, the corner where a protection trip is decided; that separation buys ≈47 of quench warning against 2.6 for the raw magnetization signal — an 18× lead time that lets the controller act on the trip rather than merely record it. A recursive Kalman/particle filter fuses the estimate across time as well as across modality, driving the state error to the full-suite floor within a few tens of milliseconds and supplying a shadow state that leads the plant. Internal state — ambipolar potential, T_e, n_e and β — is reconstructable from as few as six diagnostics at a mean coefficient of determination R^2≈0.86, confirming observability for the control layer, and a D-optimal sensor-placement criterion selects where to add or move channels to raise it. The suite adds an electron-cyclotron-emission radiometer and a fast-ion tracker on the same backbone, and the estimator is guarded by an epistemic gate: when its posterior variance or an out-of-distribution score crosses threshold, authority is handed back to a deterministic floor, with a three-detector majority vote gating every precursor trip. An in-winding nitrogen-vacancy (NV) diamond magnetometer resolves a conductor-scale quench step at a shot-noise-limited sensitivity of 0.189 pT Hz^-1/2 at 125 bandwidth, for a per-shot signal-to-noise ratio of order 10^9. The estimator holds the plasma state at the breeder design point (I_p=9.66, Q=3.076, P_ fus=85.04, δ=-0.30, centrepost peak field 16.84), which is a distinct magnet from the burner plug coil (26.49); the physical-inversion analyzers are built to the demonstrated estimator pattern on a staged, acceptance-gated schedule. Against published quench-detection, quantum-sensing and equilibrium-reconstruction benchmarks the figures are competitive in a deliberately hard, weak-signal regime.Key results (frozen anchors): L_W = 5.7 ×10^-32; Ar = 1.45 ×10^-33; fracg_B = 2 ×2.Methods & codes: FreeGSNKE, FreeGS, OpenMC, ENDF/B-VIII.0, Sauter.Live verification: 7 gate validator(s) with live-recompute cards (7 reproduced). See the Verification section and data/verification.csv.Related: Paper page · De-risking register · 3D model · Learn more about KronosPart of the 2026 Kronos publication series; independently re-run and stamped in the Kronos de-risking register (DOI 10.5281/zenodo.22645689).All numerical values are frozen design-point anchors; see the register.Public research artifact. No proprietary, financial, or supply-chain information is included.
An autonomous controller can act only on the state it can see, and in a compact spherical tokamak the state must be produced faster than the plant evolves, from a diagnostic set that survives a 14 neutron field and keeps working when individual channels drop out. We formalise the Hyperion breeder diagnostic suite — magnetics, Thomson scattering, interferometry, bolometry, and a neutron camera — and its real-time state estimator as a Bayesian inverse problem, write the forward observation model of each diagnostic and the maximum-a-posteriori estimator with its posterior covariance, and cast observability of the internal state as the rank and conditioning of the associated Fisher information. The estimator's learned core is reduced to practice at the metrics that gate it into service. A graph-neural-network reconstructor, which we identify as an amortised posterior-mean estimator over the sensor graph, recovers the plasma state and a quench precursor jointly at an area under the receiver-operating-characteristic curve (AUC) of 0.980 at a reconstruction normalised RMS error of 0.396, and degrades gracefully rather than catastrophically as channels are removed. A multi-modal fusion of a fibre-Bragg strain-rate channel with a magnetization channel reaches 0.992 (against 0.976 and 0.921 alone) and a true-positive rate of 0.857 at a 1% false-positive rate, the corner where a protection trip is decided; that separation buys ≈47 of quench warning against 2.6 for the raw magnetization signal — an 18× lead time that lets the controller act on the trip rather than merely record it. A recursive Kalman/particle filter fuses the estimate across time as well as across modality, driving the state error to the full-suite floor within a few tens of milliseconds and supplying a shadow state that leads the plant. Internal state — ambipolar potential, T_e, n_e and β — is reconstructable from as few as six diagnostics at a mean coefficient of determination R^2≈0.86, confirming observability for the control layer, and a D-optimal sensor-placement criterion selects where to add or move channels to raise it. The suite adds an electron-cyclotron-emission radiometer and a fast-ion tracker on the same backbone, and the estimator is guarded by an epistemic gate: when its posterior variance or an out-of-distribution score crosses threshold, authority is handed back to a deterministic floor, with a three-detector majority vote gating every precursor trip. An in-winding nitrogen-vacancy (NV) diamond magnetometer resolves a conductor-scale quench step at a shot-noise-limited sensitivity of 0.189 pT Hz^-1/2 at 125 bandwidth, for a per-shot signal-to-noise ratio of order 10^9. The estimator holds the plasma state at the breeder design point (I_p=9.66, Q=3.076, P_ fus=85.04, δ=-0.30, centrepost peak field 16.84), which is a distinct magnet from the burner plug coil (26.49); the physical-inversion analyzers are built to the demonstrated estimator pattern on a staged, acceptance-gated schedule. Against published quench-detection, quantum-sensing and equilibrium-reconstruction benchmarks the figures are competitive in a deliberately hard, weak-signal regime.Key results (frozen anchors): L_W = 5.7 ×10^-32; Ar = 1.45 ×10^-33; fracg_B = 2 ×2.Methods & codes: FreeGSNKE, FreeGS, OpenMC, ENDF/B-VIII.0, Sauter.Live verification: 7 gate validator(s) with live-recompute cards (7 reproduced). See the Verification section and data/verification.csv.Related: Paper page · De-risking register · 3D model · Learn more about KronosPart of the 2026 Kronos publication series; independently re-run and stamped in the Kronos de-risking register (DOI 10.5281/zenodo.22645689).All numerical values are frozen design-point anchors; see the register.Public research artifact. No proprietary, financial, or supply-chain information is included.
Priyanca Ford, P I Ford· Figshare· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
