The Reflexive Integrated Information Unit: A Differentiable Primitive for Artificial Consciousness
This work addresses the problem of creating scalable, benchmarkable components for artificial consciousness research, offering a novel computational primitive that could transform philosophical debates into empirical studies.
The paper tackles the lack of a trainable module for artificial consciousness by introducing the Reflexive Integrated Information Unit (RIIU), a recurrent cell that enhances hidden states with meta-state and broadcast buffer vectors, achieving over 90% reward restoration within 13 steps after actuator failure in a Grid-world task, twice as fast as a GRU.
Research on artificial consciousness lacks the equivalent of the perceptron: a small, trainable module that can be copied, benchmarked, and iteratively improved. We introduce the Reflexive Integrated Information Unit (RIIU), a recurrent cell that augments its hidden state $h$ with two additional vectors: (i) a meta-state $μ$ that records the cell's own causal footprint, and (ii) a broadcast buffer $B$ that exposes that footprint to the rest of the network. A sliding-window covariance and a differentiable Auto-$Φ$ surrogate let each RIIU maximize local information integration online. We prove that RIIUs (1) are end-to-end differentiable, (2) compose additively, and (3) perform $Φ$-monotone plasticity under gradient ascent. In an eight-way Grid-world, a four-layer RIIU agent restores $>90\%$ reward within 13 steps after actuator failure, twice as fast as a parameter-matched GRU, while maintaining a non-zero Auto-$Φ$ signal. By shrinking "consciousness-like" computation down to unit scale, RIIUs turn a philosophical debate into an empirical mathematical problem.