Discrete Graph Diffusion Evolution with Tabu Search for Neural Architecture Generation
Neural Architecture Generation (NAG) seeks to generate high-quality neural architectures more directly than conventional Neural Architecture Search (NAS), but existing diffusion-based approaches commonly rely on pretrained denoisers and do not naturally accommodate constrained discrete architecture spaces. This study proposes Discrete Graph Diffusion Evolution with Tabu Search (DGDETS), a framework with a training-free reverse-generation mechanism that represents architecture decisions as heterogeneous categorical variables and constructs reverse diffusion distributions directly from the current population. Fitness-weighted population members are combined with categorical diffusion compatibility to estimate clean-state distributions, from which discrete reverse transitions generate new architectures without a learned score network or graph denoiser. Constraint-aware projection repairs invalid samples while preserving posterior-supported structural decisions, and late-stage multi-start Tabu Search intensifies promising graph neighborhoods using short-term memory and aspiration. DGDETS is evaluated on NAS-Bench-101, NAS-Bench-201, NAS-Bench-301, and TransNAS-Bench-101, covering heterogeneous architecture representations and downstream tasks. The results show competitive or best-performing architecture quality across the evaluated benchmarks together with low generation time in a controlled same-hardware comparison with directly related diffusion-based NAG baselines. Ablation and sensitivity analyses further demonstrate the importance of fitness-conditioned reverse diffusion, feasibility projection, diversity preservation, and late-stage Tabu refinement. The implementation is available at https://github.com/DK-cpy/DGDETS .
Authors
- Yu Xue (ORCID: https://orcid.org/0000-0002-9069-7547)
- Pingyuan Cai
- Ferrante Neri (ORCID: https://orcid.org/0000-0002-6100-6532)
Institutions
- Nanjing University of Information Science and Technology (CN)
- Education University of Hong Kong (HK)
Publication Details
- Journal
- International Journal of Neural Systems
- Published
- 2026-11-14
- DOI
- https://doi.org/10.1142/s0129065727500341
- Primary Topic
- Evolutionary Algorithms and Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00