README.md

1# Reference-centred recurrent feedback experiment 2 3## Status and provenance 4 5This package executes a synthetic scalar experiment. It does not execute Unreal 6Engine, Pantheon, a neural model, or a validated consciousness detector. 7 8Historical source: *Consciousness_Declaration_and_Persistent_Change_Joseph_Anady.pdf*, 9Paper I, pp. 15–18. That manuscript labels its seven scalar operations as an 10illustrative executable subset, not the full historical consciousness theory. 11 12The reference-centred feedback rule below is a new proposed coupling, not a claim 13that this rule was already present in Joseph's theory or Epic Games' source. 14 15## Preserved operators 16 17f1(z)=z+1 18f2(z)=z-1 19f3(z)=z 20f4(z)=z^1 21f5(z)=1^z 22f6(z)=z/1 23f7(z)=1/z 24 25Their sequential composition is constant at one. Their parallel equal average is: 26 27F(z) = (5z + 1 + 1/z) / 7. 28 29The experiment uses this *parallel average*, not a sequential seven-stage ring. 30Its domain is finite positive z, so no reciprocal singularity is hidden. 31 32## Proposed coupling 33 34For a fixed independently specified reference r=1: 35 36G_gamma(z) = 1 + gamma [F(z) - F(1)]. 37 38The gain gamma is a dimensionless engineering parameter, not a consciousness score. 39 40Fixed points (positive-domain admissibility must be checked): 41z*=1; and z*=gamma/(5gamma-7), when the latter is defined. 42The two roots coincide at gamma=7/4. 43 44G_gamma'(1)=4gamma/7. 45 46For nonnegative gamma, the reference fixed point is locally attracting for 47gamma<7/4 and locally unstable for gamma>7/4. At equality, the derivative test is 48inconclusive; the gain value is a bifurcation boundary for this particular map. 49 50At gamma=2: 51 52G_2(z) = (10z - 5 + 2/z)/7. 53G_2(z)-z = (3z-2)(z-1)/(7z). 54G_2(z)-2/3 = (3z-2)(10z-3)/(21z). 55G_2'(2/3)=11/14. 56 57For any start in (2/3,1), G_2(z) lies strictly between 2/3 and z. The trajectory 58therefore decreases monotonically to 2/3, while the reference stays at 1. 59 60At z0=1 exactly, the state remains 1 even at gain 2. A catalyst/perturbation is 61necessary. A positive-side perturbation z0>1 grows instead; this map is not 62globally bounded. The script rejects invalid states and has a numerical guard. 63Neither a boundary clamp nor floating-point overflow counts as a result. 64 65## Reproduction 66 67Run: 68 69 python feedback_experiment.py 70 71The standard-library script verifies seven numerical checks and writes the full 72201-row trajectory (initial state plus 200 updates) to trajectory.csv. 73 74The four conditions are: 75- gain 1, initial state 0.99; 76- gain 2, initial state 0.99; 77- gain 2, exact initial state 1; 78- gain 0, initial state 0.99. 79 80## Prospective Unreal connection, not an executed integration 81 82Let U_t denote the relevant Unreal runtime state, a_t the action, r_t an 83independently defined observation/reference, and z_t the recurrent representation. 84 85a_t = pi(z_t, observations_t, memory_t) 86U_(t+1) = UnrealTick_dt(U_t,a_t) 87r_(t+1) = EncodeReference(U_(t+1)) 88z_(t+1) = r_(t+1) + gamma [F(z_t)-F(r_t)] 89 90The last rule is meaningful only on appropriately normalized positive scalar 91features, or on a separately defined vector extension. The scalar threshold 7/4 92must not be transferred to the full engine or a neural policy without measuring 93the actual closed-loop dynamics. 94 95Documented source/API entry points reviewed: 96- UWorld::Tick, Engine/Source/Runtime/Engine/Private/LevelTick.cpp 97- ULearningAgentsInteractor: observation gathering and action execution 98- ULearningAgentsPolicy: observation encoding, policy evaluation, action decoding, 99 memory state, and RunInference 100- TG_PrePhysics and TG_PostPhysics tick scheduling 101 102Official references: 103https://dev.epicgames.com/documentation/en-us/unreal-engine/API/Runtime/Engine/UWorld/Tick 104https://dev.epicgames.com/documentation/en-us/unreal-engine/API/Plugins/LearningAgents/ULearningAgentsInteractor 105https://dev.epicgames.com/documentation/en-us/unreal-engine/API/Plugins/LearningAgents/ULearningAgentsPolicy 106https://dev.epicgames.com/documentation/en-us/unreal-engine/actor-ticking-in-unreal-engine 107 108## Interpretation 109 110The exact proof and numerical run establish persistent endogenous discrepancy 111for the specified scalar recurrence and starting conditions. They do not 112establish an empirical declaration, semantic understanding, subjective experience, 113or a consciousness transition. A native integration must separately test whether 114the recurrent representation affects behavior and whether that effect survives 115appropriate matched controls. 116