Serafim Grubas | Searcharxiv

arXiv2026

The Neural Division of Labor: Biologically-Inspired Modular Architectures for Robust Neuromorphic Computing

Biological neural systems achieve high efficiency and robustness through compartmentalized architectures. In contrast, modern artificial neural networks rely on globally entangled structures, which obscure decision logic and suffer from catastrophic forgetting. Here, we report a Decomposable Spiking Neural Network (D-SNN) that eliminates global synaptic entanglement by structurally isolating classification pathways into independent experts. Optimized via a bio-inspired push-pull loss function, the D-SNN achieves competitive accuracies on MNIST, Fashion-MNIST, and CIFAR-10/100 benchmarks. This modular approach matches the performance of fully dense networks while utilizing an order of magnitude fewer parameters. In addition, our networks operate with up to several orders of magnitude lower firing rates and fewer synaptic operations. Furthermore, physically severing connections between experts provides inherent protection against catastrophic forgetting during sequential learning. Crucially, these isolated pathways generate auditable neural signals, increasing decision transparency. This biomimetic, verifiable architecture establishes an efficient foundation for deploying deterministic neuromorphic intelligence in resource-constrained edge environments.

cs.LG↗

arXiv2022

Neural Eikonal Solver: improving accuracy of physics-informed neural networks for solving eikonal equation in case of caustics

The concept of physics-informed neural networks has become a useful tool for solving differential equations due to its flexibility. There are a few approaches using this concept to solve the eikonal equation which describes the first-arrival traveltimes of acoustic and elastic waves in smooth heterogeneous velocity models. However, the challenge of the eikonal is exacerbated by the velocity models producing caustics, resulting in instabilities and deterioration of accuracy due to the non-smooth solution behaviour. In this paper, we revisit the problem of solving the eikonal equation using neural networks to tackle the caustic pathologies. We introduce the novel Neural Eikonal Solver (NES) for solving the isotropic eikonal equation in two formulations: the one-point problem is for a fixed source location; the two-point problem is for an arbitrary source-receiver pair. We present several techniques which provide stability in velocity models producing caustics: improved factorization; non-symmetric loss function based on Hamiltonian; gaussian activation; symmetrization. In our tests, NES showed the relative-mean-absolute error of about 0.2-0.4% from the second-order factored Fast Marching Method, and outperformed existing neural-network solvers giving 10-60 times lower errors and 2-30 times faster training. The inference time of NES is comparable with the Fast Marching. The one-point NES provides the most accurate solution, whereas the two-point NES provides slightly lower accuracy but gives an extremely compact representation. It can be useful in various seismic applications where massive computations are required (millions of source-receiver pairs): ray modeling, traveltime tomography, hypocenter localization, and Kirchhoff migration.

physics.geo-ph↗