Knowledge Graph-Augmented LLM Agents for Explainable Scientific Discovery and Automated Hypothesis Generation
Keywords:
knowledge graphs, large language models, scientific discovery, hypothesis generation, explainable artificial intelligence, agentic systems, provenance, research infrastructure, governanceAbstract
Scientific discovery increasingly depends on the ability to integrate vast, heterogeneous knowledge sources with computational agents capable of reasoning, generating hypotheses, and designing validation pathways. Large language models have demonstrated remarkable capacities for synthesis, abstraction, and conversational reasoning, yet their tendency toward hallucination, shallow provenance, and unstable factual grounding limits their direct use in high-stakes scientific workflows. Knowledge graphs offer structured, semantically explicit, and auditable representations of entities, relations, and evidence, making them a natural complement to generative models. This paper examines knowledge graph-augmented large language model agents as a system-level architecture for explainable scientific discovery and automated hypothesis generation. It analyzes the structural trade-offs among retrieval precision, coverage, latency, cost, and interpretability; discusses agent orchestration, provenance tracking, governance, deployment, sustainability, robustness, fairness, and policy implications; and compares applications across biomedicine, chemistry, materials science, and the social sciences. The central argument is that explainable discovery agents should not be treated as monolithic models but as socio-technical infrastructures that combine graph curation, tool use, human oversight, auditability, and reproducible evaluation. Such systems can expand the scale and diversity of hypothesis generation, but only if they are governed by transparent evidence standards, domain-specific validation, and institutional accountability. The paper concludes with forward-looking design principles for building knowledge graph-augmented agents that are scientifically useful, epistemically responsible, and operationally sustainable.
References
1. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., & Polosukhin, I. (2017). Attention is all you need. In Advances in Neural Information Processing Systems (Vol. 30). Curran Associates.
2. Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., ... & Amodei, D. (2020). Language models are few-shot learners. In Advances in Neural Information Processing Systems (Vol. 33, pp. 1877-1901).
3. Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., ... & Zhou, D. (2022). Chain-of-thought prompting elicits reasoning in large language models. In Advances in Neural Information Processing Systems (Vol. 35, pp. 24824-24837).
4. Bubeck, S., Chandrasekaran, V., Eldan, R., Gehrke, J., Horvitz, E., Kamar, E., ... & Zhang, Y. (2023). Sparks of artificial general intelligence: Early experiments with GPT-4. arXiv preprint arXiv:2303.12712.
5. Jiang, Hao, et al. "AgentOmnia: Scaling Agentic Models for Full-Scenario Applications." arXiv preprint arXiv:2607.23124 (2026).
6. Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., ... & Fung, P. (2023). Survey of hallucination in natural language generation. ACM Computing Surveys, 55(12), 1-38.
7. Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., ... & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive NLP tasks. In Advances in Neural Information Processing Systems (Vol. 33, pp. 9459-9474).
8. Hogan, A., Blomqvist, E., Cochez, M., d'Amato, C., de Melo, G., Gutierrez, C., ... & Zimmermann, A. (2021). Knowledge graphs. ACM Computing Surveys, 54(4), 1-37.
9. Paulheim, H. (2017). Knowledge graph refinement: A survey of approaches and evaluation methods. Semantic Web, 8(3), 489-508.
10. King, R. D., Rowland, J., Oliver, S. G., Young, M., Aubrey, W., Byrne, E., ... & Whelan, K. E. (2009). The automation of science. Science, 324(5923), 85-89.
11. Schmidt, M., & Lipson, H. (2009). Distilling free-form natural laws from experimental data. Science, 324(5923), 81-85.
12. Wilkinson, M. D., Dumontier, M., Aalbersberg, I. J., Appleton, G., Axton, M., Baak, A., ... & Mons, B. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data, 3, 160018.
13. Topol, E. J. (2019). High-performance medicine: The convergence of human and artificial intelligence. Nature Medicine, 25(1), 44-56.
14. Mitchell, M., Wu, S., Zaldivar, A., Barnes, P., Vasserman, L., Hutchinson, B., ... & Gebru, T. (2019). Model cards for model reporting. In Proceedings of the Conference on Fairness, Accountability, and Transparency (pp. 220-229). ACM.
15. Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., Daumé III, H., & Crawford, K. (2021). Datasheets for datasets. Communications of the ACM, 64(12), 86-92.
16. Raji, I. D., Smart, A., White, R. N., Mitchell, M., Gebru, T., Hutchinson, B., ... & Barnes, P. (2020). Closing the AI accountability gap: Defining an end-to-end framework for internal algorithmic auditing. In Proceedings of the Conference on Fairness, Accountability, and Transparency (pp. 33-44). ACM.
17. Shneiderman, B. (2020). Human-centered AI. Oxford University Press.
18. Amodei, D., Olah, C., Steinhardt, J., Christiano, P., Schulman, J., & Mané, D. (2016). Concrete problems in AI safety. arXiv preprint arXiv:1606.06565.
19. Marcus, G. (2018). Deep learning: A critical appraisal. arXiv preprint arXiv:1801.00631.
20. Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., ... & Liang, P. (2021). On the opportunities and risks of foundation models. arXiv preprint arXiv:2108.07258.
21. Wang, L., Ma, C., Feng, X., Zhang, Z., Yang, H., Zhang, J., ... & Wen, J. R. (2024). A survey on large language model based autonomous agents. Frontiers of Computer Science, 18(6), 186345.
22. Boiko, D. A., MacKnight, R., Kline, B., & Gomes, G. (2023). Autonomous chemical research with large language models. Nature, 624(7992), 570-578.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 International Journal of Artificial Intelligence Engineering and Systems
This work is licensed under a Creative Commons Attribution 4.0 International License.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
