Quantum Energy Packet Structures

Building on QEP theory (space as cells, everything as quantum energy packets), this preprint describes every fundamental fermion as a bundle of three QEP strands. Strand twist gives the charge, the axis of the odd strand gives the color, the axis of QEP circulation gives the generation, and the tether to the cell network gives spin ½. The charge and color counting coincides with the Harari–Shupe and Bilson-Thompson models, and this is acknowledged. An exact linear analysis of 51 reactions shows that the number of QEPs is not conserved; only the counts of twisted and untwisted strands are, which correspond to electric charge, baryon number and lepton number (the three separate lepton numbers hold in these reactions but are broken by neutrino oscillations, which the strand picture allows). For hadron masses we construct a phenomenological model in which mass comes from the organization of the bundles. Its ingredients are relativistic kinetic energy of Gaussian bundles, a Y-shaped string with Steiner length, short-range attraction, a smeared spin–spin term and a node energy. With 13 parameters the model fits 42 hadrons with a mean deviation of 12.4 MeV (0.40%) and a maximum of 1.17%. The paper tests the model in several ways: Cross-validation: in leave-one-out cross-validation the mean deviation is 0.64%, against 1.57% (fit) and 2.03% (cross-validation) for a standard De Rújula–Georgi–Glashow-type quark model on the same data. Blind test: three baryons discovered in 2012–2017 were withheld and reproduced within 0.4–0.6%. 2026 measurements: comparisons with Ξcc⁺, Ωcc⁺ and Bc* are reported honestly as post-hoc checks, not as predictions. Uncertainties and limitations: parameter uncertainties and correlations are given, and a known weakness is documented and propagated: hyperfine splittings in charm and doubly heavy mesons are too small. Predictions for undiscovered states: Coreon (Ξbc): 6947 ± 30 MeV, spin ½, lifetime 0.38–0.51 ps, with its production conditions and decay channels. Alignon (Ωb*⁻): 6071 ± 5 MeV; it decays only by emitting a ≈25 MeV photon. Ωccc and Ωbbb: 4879 and 14241 MeV. These are the most distinctive predictions, since they differ from lattice QCD by +83 and −125 MeV. All predictions are compared with lattice QCD and other model calculations, and each has an explicit falsification condition. For leptons, the Koide relation is read through the three strand axes, giving the tau mass to 0.002%. Its application to neutrinos follows Brannen: normal ordering, an almost massless lightest neutrino, and a mass sum of 59.3 meV. A reproducibility package included in this record (data for 42 hadrons, parameter sets of the final, pre-2026 and pre-2012 fits, and code for the mass model, cross-validation, uncertainty analysis and figures) accompanies the paper. This record also serves to timestamp the Coreon and Alignon predictions before any measurement. The use of an AI assistant for calculations, figures and editing is disclosed in the paper.