ATRUVAI research
Public research notes from the Zarvok mathematical architecture, split into readable articles. No download wall, no version label, no internal deployment coordinates.
Runtime mathematics
A private AI individual is treated as a measurable runtime object, not a display name. Its state includes identity, sandbox, memory, regulation, world model, context binding, privacy state, chronology, and microcircuit telemetry.
Read noteIdentity-time
Identity-time continuity is represented as a hash-linked sequence of private runtime events. A copied state can be detected as lineage conflict without claiming physical non-clonability.
Read noteControl field
The regulatory field adjusts salience, threshold, gain, inhibition, and memory-pass pressure. These are computational control variables, not claims of emotion or consciousness.
Read noteMemory and adaptation
Memory separation prevents similar names or contexts from collapsing into one private trace. Fluid control lets runtime rules adapt while bounded NoGo pressure suppresses irrelevant learned rules.
Read noteWorld and context
Zarvok's public mathematical frame separates active state from counterfactual world traces and binds object, place, time, relation, and target into contextual hashes.
Read notePrivacy and field separation
Private coordinate frames and access-control calculus separate individuals under the same visible label and gate access to sealed private records.
Read noteProof surface
The proof frame combines identity separation, timeline validity, regulatory telemetry, memory separation, fluid stability, world traces, context binding, privacy enforcement, phase divergence, and lineage control.
Read note