Search papers, labs, and topics across Lattice.
This study explores the use of a Coherent Ising Machine (CIM) to train energy-based neural networks through Equilibrium Propagation, achieving performance on par with traditional software methods. By integrating the Adam optimizer to refine the ground state of a Hopfield energy network, the researchers significantly enhance both convergence speed and solution accuracy. The findings indicate that CIM dynamics can effectively scale to deeper architectures and convolutional operations, paving the way for energy-efficient AI hardware solutions.
CIM dynamics could revolutionize neural network training, achieving efficiency and scalability that traditional methods struggle to match.
While Ising machines serve as advanced physical solvers for the Ising model,enabling applications in combinatorial optimization and neural network training,their scalability for large-scale neural networks remains constrained by hardware connectivity limitations and suboptimal training methodologies. In this work,we leverage a Coherent Ising Machine (CIM) to train an energy-based neural network using Equilibrium Propagation, achieving performance comparable to existing software-based implementations. We further enhance the algorithm by integrating the Adam optimizer to solve for the ground state of a Hopfield energy network, significantly improving convergence speed and solution accuracy. Additionally, we demonstrate the scalability of our approach across deeper network architectures and convolutional operations. Our results highlight the potential of CIM dynamics as a scalable platform for training complex neural networks, offering a pathway toward energy-efficient implementations via analog circuits, optoelectronics, or integrated photonics. This work establishes a novel physical framework for next-generation AI hardware development.
