languageModelAlignmentFraction
plain-language theorem explainer
Language model alignment fraction is the constant rational 7/8. Researchers auditing Recognition Science empirical checks cite it when summarizing the language-model substrate result inside the three-substrate certificate. The declaration is a bare constant assignment with no computation or lemmas.
Claim. The language model alignment fraction equals $7/8$.
background
The ThreeSubstrateValidationCert module assembles J-cost checks across language models, photonic qubits, and magnetized plasma. J-cost is the function satisfying J(1) = 0, J(r) = J(r^{-1}), and positive descent away from the fixed point. The module imports Cost and MultiChannelJCost to access these properties and records the language-model result as the fraction 7/8.
proof idea
Direct constant definition with empty body.
why it matters
Supplies the lm_alignment field of the ThreeSubstrateCert structure and is unfolded in lm_fraction_eq, lm_above_threshold, and seven_eighths_from_F2_cube. The value 7/8 = (2^3 - 1)/2^3 connects to the eight-tick octave of the forcing chain. The module states the entire certificate remains at hypothesis grade.
Switch to Lean above to see the machine-checked source, dependencies, and usage graph.