A self-evolving neural architecture search framework inspired by NCA, in which a population of NCA-based controllers autonomously generates and improves candidate GNN architectures over successive generations, enabling progressive adaptation of the architecture distribution without hand-crafted design rules.
Abstract
Designing effective graph neural networks for node classification still depends heavily on human intuition, repeated trial-and-error, and expert knowledge of architectural components. To reduce this manual burden, we introduce a self-evolving neural architecture search framework inspired by Neural Cellular Automata (NCA), in which a population of NCA-based controllers autonomously generates and improves candidate GNN architectures over successive generations. Each controller encodes layer-wise architectural decisions, including layer activation, message-passing operator type, and hidden dimensionality, thereby defining a dynamic search space of graph models. The sampled architectures are trained and evaluated on node classification tasks, while their quality is assessed through a fitness function that jointly considers validation accuracy, validation loss, architectural depth, and parameter count relative to a baseline model. In this way, the proposed framework promotes not only predictive performance but also structural efficiency. Evolution proceeds through elitist retention, tournament selection, and mutation of controller parameters, enabling progressive adaptation of the architecture distribution without hand-crafted design rules.
A probabilistic L-system-inspired encoding for architecture search, where each individual is a small set of production rules and repeated rewriting generates a symbol sequence that is translated into a GNN.
Maciej Krzywda, Szymon Łukasik, Amir H. Gandomi· GECCO Companion· 0 citations
A pairwise surrogate model is introduced that predicts whether an offspring architecture is likely to outperform its parent using genotype-derived features and early training signals, enabling full evaluation only for promising candidates within the evolutionary loop.
Maciej Krzywda, Szymon Łukasik, Amir H. Gandomi· GECCO Companion· 0 citations
Self-organization is an emergent property of life, driven by the collective behavior of individual components acting on local information. Biological neurons, through local interactions transmitted through synapses, are able to learn efficiently and can adapt their connections over an organism's lifespan. Motivated by these desirable properties of adaptability and local interaction, neural cellular automata (NCA) models have been successful at learning morphogenesis solely through local update rules, demonstrating stability over many updates and robustness to perturbations. In this work, we introduce Meta Neural Cellular Automata (MetaNCA), a framework that learns local rules which self-organize the weights of artificial neural networks. A learned rule network iteratively updates the weights of a task network using only local interactions on the computation graph. We propose a novel Weight Transformer architecture for the local rule network, which uses linear attention to aggregate signals from neighboring weights and hidden states. Once trained, the rule network generates task networks of diverse architectures without backpropagation. We show that MetaNCA generates weights for feedforward MLPs, CNNs, and ResNets on MNIST and CIFAR-100, scaling to networks of 2 million parameters. We further show that MetaNCA generalizes to architectures not seen during meta-training, and that architectural diversity in the training phase strengthens this generalization.
Meet Barot, Daniel Berenberg, Sina Khajehabdollahi· 0 citations
Network optimization driven by Artificial Intelligence (AI) is increasingly required to operate under non-stationary environments, where traffic conditions exhibit cyclic patterns as well as abrupt structural changes. Most existing learning-based approaches rely on explicit retraining or change detection, which limits autonomous adaptation. This paper proposes a memory-enabled AI framework for network optimization based on Quantum Gene Regulatory Networks (QGRNs), a biologically inspired model that probabilistically encodes and recalls multiple network control strategies. By combining the QGRNs with a genetic algorithm, the proposed method realizes two complementary adaptation mechanisms: short-term epigenetic-like adaptation to cyclic fluctuations and long-term evolutionary adaptation to structural shifts without requiring explicit environmental change detection. We apply the AI framework to a network slice embedding problem in dynamic edge–cloud environments. Simulation results show that, once evolutional mechanism finds and memorizes the optimal configurations after structural changes, the proposed method recalls them autonomously and adapts faster and more stably than conventional GA-, GRN-, and bandit-based approaches. These results suggest that memory-enabled models such as QGRNs provide a foundation for AI-driven network optimization under persistent non-stationarity.
Kazuki Sekizawa, Masaaki Yamauchi, D. Kominami et al.· International Conference on...· 0 citations
Evolutionary program search guided by Large Language Models (LLMs) has emerged as a powerful paradigm for automated scientific discovery. However, current approaches are fundamentally constrained by three bottlenecks: structurally blind parent selection, sparse whole-program evaluation rewards, and static mutation operators that fail to adapt during search. We present GAE (Graph-Augmented Evolution), a framework that resolves these limitations through a tightly coupled, three-pillar architecture. First, a relational graph neural network (GNN) parses programs into typed computation graphs, producing structure-aware embeddings. Second, an RL-optimized meta-controller leverages these embeddings to replace blind evolutionary sampling with a directed policy, dynamically selecting optimal parents and mutation directions based on reward history. Third, an online GRPO fine-tuning loop continuously updates the LLM mutation operator at test-time using group-normalized evaluation rewards, directly aligning the model's generation distribution with high-fitness structural edits. We evaluate GAE on a challenging scientific discovery task: symbolic regression for complex nonlinear oscillator systems. By transforming stochastic search into a directed, self-improving trajectory, GAE efficiently discovers closed-form physical equations, consistently matching or outperforming static LLM-driven baselines and achieving state-of-the-art out-of-distribution performance.
Ensemble-GNAS is proposed, a novel framework that integrates evolutionary GNAS with ensemble learning to address limitations of existing graph NAS algorithms and outperforms the state-of theart methods in both classification accuracy and performance stability.
Weifeng Guo, Pengyu Wang, Ying Bi et al.· IEEE Transactions on Neural...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
