EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Advancing Model Transparency via Self-Explaining Deep Learning Architectures Integrating Symbolic Logic and Neural Representations

Timothy Langford

International Journal of Artificial Intelligence Research · 19.05.2026 · 1 (2)

DOI: 10.66280/ijair.v1i2.138 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Artificial Intelligence Research, online erschienen am 19.05.2026. Verfasst von Timothy Langford. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The rapid integration of deep learning architectures into critical societal infrastructures has highlighted a fundamental tension between predictive efficacy and structural interpretability. While contemporary neural networks exhibit unprecedented performance in high-dimensional pattern recognition, their inherent opacity poses substantial risks to accountability, safety, and institutional trust. This paper explores the advancement of model transparency through the development of self-explaining architectures that natively integrate symbolic logic with connectionist neural representations. Unlike post-hoc interpretability methods that provide approximate justifications for black-box decisions, self-explaining systems aim to produce human-legible reasoning as an intrinsic component of the computational process. By embedding logical constraints and symbolic abstractions directly into the neural fabric, these hybrid systems offer a robust pathway toward verifiable and auditable artificial intelligence. The research provides a comprehensive system-level analysis of the structural trade-offs involved in hybridizing logic and deep learning, with a specific focus on architecture, governance, and deployment sustainability. Furthermore, the discussion extends to the socio-technical implications of these architectures, examining how integrated transparency affects fairness, policy compliance, and the long-term robustness of critical decision systems. Through detailed conceptual analysis and cross-domain comparisons, this work argues that the convergence of symbolic reasoning and neural learning is not merely a technical improvement but a necessary evolution for the responsible deployment of large-scale intelligent infrastructures.

Zitieren

APA

Langford, T. (2026). Advancing Model Transparency via Self-Explaining Deep Learning Architectures Integrating Symbolic Logic and Neural Representations. International Journal of Artificial Intelligence Research, 1(2). https://doi.org/10.66280/ijair.v1i2.138

Vancouver

Langford T. Advancing Model Transparency via Self-Explaining Deep Learning Architectures Integrating Symbolic Logic and Neural Representations. International Journal of Artificial Intelligence Research. 2026;1(2). doi:10.66280/ijair.v1i2.138

BibTeX
@article{Langford2026,
  author = {Langford, Timothy},
  title = {Advancing Model Transparency via Self-Explaining Deep Learning Architectures Integrating Symbolic Logic and Neural Representations},
  journal = {International Journal of Artificial Intelligence Research},
  year = {2026},
  volume = {1},
  number = {2},
  doi = {10.66280/ijair.v1i2.138}
}
Kostenlos starten →