Triad Logic as Neural Constraint: A Mobile-Friendly Toy Simulation of Identity, Distinction, and Binding

A TL-based brain model with measurable Identity, Distinction, and Binding failures. Copy/paste-friendly Description: This paper proposes a falsifiable bridge between Triad Logic (TL) and computational neuroscience by treating TL not as a claim that neurons “compute logic,” but as a set of structural constraints that any stable meaning-bearing dynamical system must satisfy. We implement a leaky-integrator recurrent network with three tunable control parameters: recurrent excitation κW (Identity), lateral inhibition γ (Distinction), and a goal-driven eligibility trace update rate λtr (Binding). Each parameter is paired with an operational diagnostic: post-cue retention failure Ifail, decision clarity Dpmarg (softmax probability gap), and active-without-trace fragmentation Bfrag. To validate that the three TL legs are experimentally separable in the toy system, the paper includes stress-test sweeps and a correlation-based Separation Index between each knob and its target metric. A complete mobile-friendly Python reference implementation (tested on Pydroid) is provided to make the model easy to run, modify, and reproduce. The result is a practical computational sandbox for exploring “computational phenotypes” corresponding to drift (Identity failure), collapse/blur (Distinction failure), and fragmentation/intrusion (Binding failure).