LIE: A Law Identification Engine for Heterogeneous Trajectories

LIE: A Law Identification Engine for Heterogeneous Trajectories The Law Identification Engine (LIE) is a human-directed computational research instrument developed to investigate whether heterogeneous observed trajectories arise from shared governing rules or from distinguishable alternatives. The work is motivated by the Activation–Constraint–Tissue Context (ACT) framework and the Permissible Phenotype Conjecture (PPC), which proposes a hierarchy from invariant governing rules (L), through state-dependent dynamics (F) and permissible futures (V), to realized trajectories (γ). LIE addresses a corresponding inverse problem: given incomplete trajectories, constrained host variation, and competing explanations, can an investigator distinguish the governing structures responsible? Developmental evaluation LIE v0.9 was evaluated in a frozen synthetic campaign comprising 352 near-different-law and 120 shared-law lineages at nested evidence depths of 24 and 48 four-member families. Sixteen diagnostic controls brought the total to 960 completed study rungs. Guarded correct identification of near-different laws increased from 15.1% to 55.4% with additional family evidence, a 40.34-percentage-point improvement. Guarded recognition of genuinely shared laws reached 96.7% at 48 families. The experiment also preserved unresolved cases and high-confidence errors. Twelve incorrect shared-law declarations retained posterior sharing probabilities of at least 0.99, illustrating that confidence does not guarantee structural identification. Contents and reproducibility This archive contains the frozen scientific implementation, Campaign 2.0 configuration and seeds, numerical reference inputs, validation tests, administrative repair provenance, the research manuscript, supplementary methods, and compact aggregate results with arithmetic-audit tools. Large generated observations, posterior arrays, and original case-level production analysis exports are excluded. The supplied code supports inspection and fresh synthetic regeneration; the compact summaries do not independently reproduce every production analysis. Scope This is a developmental synthetic evaluation, not unrestricted law discovery, a validated biological model, an independently blinded final qualification, or a clinical decision-making instrument. Its scientific objective is to make candidate governing explanations subject to explicit comparison, prediction, uncertainty, and falsification. Research conducted independently, without Accenture funding, sponsorship, or endorsement.