Paper proposes five-layer implementation architecture for S3Q theory of machine qualia
An arXiv preprint (arXiv:2609.30743v1) proposes a five-layer implementation architecture that maps the S3Q (Simulated, Situated, Structurally Coherent) theory of consciousness to compatible computational primitives, operating on continuous, differentiable per-object slot vectors. The paper composes existing machine-learning components into a single pipeline, outlines a developmental bootstrap sequence, and claims individually falsifiable predictions about emergent self-like behavior and three behavioral patterns (hesitation, curiosity, avoidance) tied to prediction errors.