oscillator-networks

A coupled oscillator field where the local learning rules (α_i, β_i, γ_i) themselves evolve as a second-order dynamical system, constrained by an identity curvature functional H[L], with Lyapunov-gated rollback of destructive meta-updates — unified in a single implementable architecture.

Discreteness as the Late-Time Limit of a Smooth Flow

Universal foundational architecture specification for the Physical Matrix Neural Network (PMNN) superseding the Transformer paradigm.

budgeted hebbian kuramoto with fixed support and symmetric-frobenius projection — control/calibration algorithm for oscillator-based ising machines