Agentic deep graph reasoning yields self-organizing knowledge networks
Source abstract
Abstract
Abstract We present an agentic, autonomous graph expansion framework that iteratively structures and refines knowledge in situ. Unlike conventional knowledge graph construction methods relying on static extraction or single-pass learning, our approach couples a reasoning-native large language model with a continually updated graph representation. At each step, the system actively generates new concepts and relationships, merges them into a global graph, and formulates subsequent prompts based on its evolving structure. Through this feedback-driven loop, the model organizes information into a scale-free network characterized by hub formation, stable modularity, and bridging nodes that link disparate knowledge clusters. Over hundreds of iterations, new nodes and edges continue to appear without saturating, while centrality measures and shortest path distributions evolve to yield increasingly distributed connectivity. Applied to materials design problems, we present compositional reasoning experiments to foster knowledge synthesis, yielding cross-domain ideas that transcend rote summarization. Graphic abstract
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Machine Learning in Materials Science · Advanced Graph Neural Networks · Topic Modeling
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (62)
Show source-provided references
1. T.S. Kuhn, The Structure of Scientific Revolutions (University of Chicago Press, 1962)
2. D. Spivak, T. Giesa, E. Wood, M. Buehler, Category theoretic analysis of hierarchical protein materials and social networks. PLoS ONE 6, e23911 (2011)10.1371/journal.pone.0023911 ↗
3. T. Giesa, D. Spivak, M. Buehler, Reoccurring patterns in hierarchical protein materials and music: the power of analogies. BioNanoScience 1, 153 (2011)10.1007/s12668-011-0022-5 ↗
4. T. Giesa, D. Spivak, M. Buehler, Category theory based solution for the building block replacement problem in materials design. Adv. Eng. Mater. 14, 810 (2012)10.1002/adem.201200109 ↗
5. A. Vaswani et al., Attention is all you need (2017). https://papers.nips.cc/paper/7181-attention-is-all-you-need
6. A. Radford, K. Narasimhan, T. Salimans, I. Sutskever, Improving language understanding by generative pre-training https://gluebenchmark.com/leaderboard
7. L. Xue et al., ByT5: towards a token-free future with pre-trained byte-to-byte models. Trans. Assoc. Comput. Linguist. 10, 291–306 (2021). arXiv:2105.13626v310.1162/tacl_a_00461 ↗
8. A.Q. Jiang et al., Mistral 7B (2023). arXiv:2310.06825
9. Phi-2: The surprising power of small language models - Microsoft research. https://www.microsoft.com/en-us/research/blog/phi-2-the-surprising-power-of-small-language-models/
10. A. Dubey et al., The llama 3 herd of models (2024). arXiv:2407.21783
11. T.B. Brown et al., Language models are few-shot learners (2020)
12. H. Salinas et al., Exoplanet transit candidate identification in Tess full-frame images via a transformer-based algorithm (2025). arXiv:2502.0754210.1093/mnras/staf347 ↗
13. J. Schmidt, M.R.G. Marques, S. Botti, M.A.L. Marques, Recent advances and applications of machine learning in solid-state materials science. NPJ Comput. Mater. (2019). https://doi.org/10.1038/s41524-019-0221-010.1038/s41524-019-0221-0 ↗
14. E.L. Buehler, M.J. Buehler, X-LoRA: Mixture of low-rank adapter experts, a flexible framework for large language models with applications in protein mechanics and design. APL Mach. Learn. 2, 026119 (2024)10.1063/5.0203126 ↗
15. S.E. Arevalo, M.J. Buehler, Learning from nature by leveraging integrative biomateriomics modeling toward adaptive and functional materials. MRS Bull. (2023). https://doi.org/10.1557/s43577-023-00610-810.1557/s43577-023-00610-8 ↗
16. Y. Hu, M.J. Buehler, Deep language models for interpretative and predictive materials science. APL Mach. Learn. 1, 010901 (2023). https://doi.org/10.1063/5.013431710.1063/5.0134317 ↗
17. N.J. Szymanski et al., Toward autonomous design and synthesis of novel inorganic materials. Mater. Horiz. 8, 2169–2198 (2021). https://doi.org/10.1039/D1MH00495F10.1039/D1MH00495F ↗
18. J. Vamathevan et al., Applications of machine learning in drug discovery and development. Nat. Rev. Drug Discovery 18, 463–477 (2019)10.1038/s41573-019-0024-5 ↗
19. J. Jumper et al., Highly accurate protein structure prediction with AlphaFold. Nature 596, 1–12 (2021)10.1038/s41586-021-03819-2 ↗
20. Protein structure prediction by trRosetta. https://yanglab.nankai.edu.cn/trRosetta/
21. R. Wu et al., High-resolution de novo structure prediction from primary sequence. bioRxiv 2022.07.21.500999 (2022). https://www.biorxiv.org/content/10.1101/2022.07.21.500999v110.1101/2022.07.21.500999 ↗
22. V. Abbott, G. Zardini, Flashattention on a napkin: a diagrammatic approach to deep learning io-awareness (2024). arXiv:2412.03317
23. M.J. Buehler, Graph-aware isomorphic attention for adaptive dynamics in transformers. APL Mach. Learn. 3, 026108 (2025). https://doi.org/10.1063/5.025687310.1063/5.0256873 ↗
24. T. Miconi, K. Kay, Neural mechanisms of relational learning and fast knowledge reassembly in plastic neural networks. Nat. Neurosci. 28, 406–414 (2025)10.1038/s41593-024-01852-8 ↗
25. OpenAI et al., OpenAI o1 system card (2024). arXiv:2412.16720
26. M.J. Buehler, PRefLexOR: Preference-based recursive language modeling for exploratory optimization of reasoning and agentic thinking. npj Artif. Intell. 1, Article 4 (2025). https://doi.org/10.1038/s44230-025-00005-910.1038/s44230-025-00005-9 ↗
27. M.J. Buehler, In situ graph reasoning and knowledge expansion using Graph-PRefLexOR. Adv. Intell. Discov. 202500006 (2025). https://doi.org/10.1002/aidi.20250000610.1002/aidi.202500006 ↗
28. C.K. Reddy, P. Shojaee, Towards scientific discovery with generative ai: progress, opportunities, and challenges (2024). arXiv preprint arXiv:2412.11427
29. M.J. Buehler, Accelerating scientific discovery with generative knowledge extraction, graph-based representation, and multimodal intelligent graph reasoning. Mach. Learn.: Sci. Technol. 5, 035083 (2024).
30. S. Brin, Extracting patterns and relations from the world wide web. In International Workshop on the World Wide Web and Databases (WebDB), 172–183 (1998)10.1007/10704656_11 ↗
31. O. Etzioni et al., Knowitall: Fast, scalable, and self-supervised web information extraction. In Proceedings of the 13th International World Wide Web Conference (WWW), 100–110 (2004)10.1145/988672.988687 ↗
32. M. Banko, M.J. Cafarella, S. Soderland, M. Broadhead, O. Etzioni, Open information extraction from the web. In Proceedings of the 20th International Joint Conference on Artificial Intelligence (IJCAI), 2670–2676 (2007)
33. O. Etzioni, A. Fader, J. Christensen, S. Soderland, Mausam, Open information extraction: The second generation. In Proceedings of the 22nd International Joint Conference on Artificial Intelligence (IJCAI), 3–10 (2011)
34. A. Bordes, N. Usunier, A. Garcia-Duran, J. Weston, O. Yakhnenko, Translating embeddings for modeling multi-relational data. In Advances in Neural Information Processing Systems (NeurIPS), 2787–2795 (2013)
35. L.A. Galárraga, C. Teflioudi, K. Hose, F.M. Suchanek, Amie: Association rule mining under incomplete evidence in ontological knowledge bases. In Proceedings of the 22nd International World Wide Web Conference (WWW), 413–422 (2013)10.1145/2488388.2488425 ↗
36. A. Carlson et al., Toward an architecture for never-ending language learning. In Proceedings of the 24th AAAI Conference on Artificial Intelligence (AAAI), 1306–1313 (2010)10.1609/aaai.v24i1.7519 ↗
37. X.L. Dong et al., Knowledge vault: A web-scale approach to probabilistic knowledge fusion. In Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 601–610 (2014)10.1145/2623330.2623623 ↗
38. W. Xiong, T. Hoang, W.Y. Wang, Deeppath: A reinforcement learning method for knowledge graph reasoning. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing (EMNLP), 564–573 (2017)10.18653/v1/D17-1060 ↗
39. D.R. Swanson, Fish oil, Raynaud’s syndrome, and undiscovered public knowledge. Perspect. Biol. Med. 30, 7–18 (1986)10.1353/pbm.1986.0087 ↗
40. D. Cameron et al., A graph-based recovery and decomposition of Swanson’s hypothesis using semantic predications. J. Biomed. Inform. 46, 238–251 (2013). https://doi.org/10.1016/j.jbi.2012.09.00410.1016/j.jbi.2012.09.004 ↗
41. M. Nickel, K. Murphy, V. Tresp, E. Gabrilovich, A review of relational machine learning for knowledge graphs. Proc. IEEE 104, 11–33 (2016)10.1109/JPROC.2015.2483592 ↗
42. A.-L. Barabási, R. Albert, Emergence of scaling in random networks. Science 286, 509–512 (1999)10.1126/science.286.5439.509 ↗
43. S.R. White et al., Autonomic healing of polymer composites. Nature 409, 794–797 (2001)10.1038/35057232 ↗
44. Y. Bar-Yam, Dynamics of complex systems ISBN 0813341213 (1997). https://necsi.edu/dynamics-of-complex-systems
45. B. Bhushan, Biomimetics: lessons from nature–an overview. Phil. Trans. R. Soc. A 367, 1445–1486 (2009)10.1098/rsta.2009.0011 ↗
46. D. Nepal et al., Hierarchically structured bioinspired nanocomposites. Nat. Mater. 2022, 1–18 (2022)
47. J.A. Fodor, Z.W. Pylyshyn, Connectionism and cognitive architecture: a critical analysis. Cognition 28, 3–71 (1988)10.1016/0010-0277(88)90031-5 ↗
48. J. Zhao et al., Exploring the compositional deficiency of large language models in mathematical reasoning. arXiv preprint arXiv:2405.06680 (2024)
49. J. Shi et al., Cryptox: Compositional reasoning evaluation of large language models. arXiv preprint arXiv:2502.07813 (2025)
50. Z. Xu, Z. Shi, Y. Liang, Do large language models have compositional ability? An investigation into limitations and scalability (2024). arXiv:2407.15720
51. A. Ghafarollahi, M.J. Buehler, SciAgents: automating scientific discovery through bioinspired multi-agent intelligent graph reasoning. Adv. Mater. 37, 2413523 (2025). https://doi.org/10.1002/adma.20241352310.1002/adma.202413523 ↗
52. OpenAI. OpenAI o3-mini system card (2025). https://openai.com/index/o3-mini-system-card/
53. J. Geiping et al., Scaling up test-time compute with latent reasoning: a recurrent depth approach (2025). arXiv:2502.05171
54. S. Arevalo, M.J. Buehler, Learning from nature by leveraging integrative biomateriomics modeling toward adaptive and functional materials. MRS Bull. (2023). https://doi.org/10.1557/s43577-023-00610-810.1557/s43577-023-00610-8 ↗
55. V. Tshitoyan et al., Unsupervised word embeddings capture latent knowledge from materials science literature. Nature 571, 95–98 (2019)10.1038/s41586-019-1335-8 ↗
56. M.J. Buehler, Generating 3D architectured nature-inspired materials and granular media using diffusion models based on language cues. Oxford Open Materials Science (2022). https://academic.oup.com/ooms/article/2/1/itac010/682354210.1093/oxfmat/itac010 ↗
57. M.J. Buehler, Predicting mechanical fields near cracks using a progressive transformer diffusion model and exploration of generalization capacity. J. Mater. Res. 38, 1317–1331 (2023). https://doi.org/10.1557/s43578-023-00892-310.1557/s43578-023-00892-3 ↗
58. L.C. Brinson et al., Community action on FAIR data will fuel a revolution in materials research. MRS Bull. (2023). https://doi.org/10.1557/s43577-023-00498-410.1557/s43577-023-00498-4 ↗
59. E. Stach et al., Autonomous experimentation systems for materials development: a community perspective. Matter 4, 2702–2726 (2021)10.1016/j.matt.2021.06.036 ↗
60. networkx/networkx: Network Analysis in Python. https://github.com/networkx/networkx
61. M. Bastian, S. Heymann, M. Jacomy, Gephi: An open source software for exploring and manipulating networks (2009). http://www.aaai.org/ocs/index.php/ICWSM/09/paper/view/15410.1609/icwsm.v3i1.13937 ↗
62. P. Shannon et al., Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 13, 2498–2504 (2003)10.1101/gr.1239303 ↗