{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T23:01:49Z","timestamp":1789686109817,"version":"4.0.1"},"reference-count":67,"publisher":"Springer Science and Business Media LLC","issue":"9","license":[{"start":{"date-parts":[[2026,9,10]],"date-time":"2026-09-10T00:00:00Z","timestamp":1788998400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"},{"start":{"date-parts":[[2026,9,10]],"date-time":"2026-09-10T00:00:00Z","timestamp":1788998400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"}],"funder":[{"DOI":"10.13039\/501100002858","name":"China Postdoctoral Science Foundation","doi-asserted-by":"publisher","award":["2025M780875"],"award-info":[{"award-number":["2025M780875"]}],"id":[{"id":"10.13039\/501100002858","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Nat Mach Intell"],"DOI":"10.1038\/s42256-026-01298-6","type":"journal-article","created":{"date-parts":[[2026,9,10]],"date-time":"2026-09-10T09:03:51Z","timestamp":1789031031000},"page":"1441-1456","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["A collaborative agent with two lightweight synergistic models for autonomous crystal materials research"],"prefix":"10.1038","volume":"8","author":[{"given":"Tongyu","family":"Shi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yutang","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhanyuan","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qian","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jie","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wenhe","family":"Xu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yang","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8431-4431","authenticated-orcid":false,"given":"Dawei","family":"Dai","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6962-284X","authenticated-orcid":false,"given":"Rui","family":"He","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wenhua","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6743-7923","authenticated-orcid":false,"given":"Jiahong","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2566-6194","authenticated-orcid":false,"given":"Xue-Feng","family":"Yu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,9,10]]},"reference":[{"key":"1298_CR1","doi-asserted-by":"publisher","first-page":"331","DOI":"10.1038\/s41586-025-09516-8","volume":"646","author":"F Hinrichs","year":"2025","unstructured":"Hinrichs, F. et al. A ductile chromium\u2013molybdenum alloy resistant to high-temperature oxidation. Nature 646, 331\u2013337 (2025).","journal-title":"Nature"},{"key":"1298_CR2","doi-asserted-by":"publisher","first-page":"142","DOI":"10.1038\/s44286-025-00186-x","volume":"2","author":"Z Fang","year":"2025","unstructured":"Fang, Z., Zhu, P., Zhang, X., Feng, Y. & Wang, H. Self-looped electrochemical recycling of lithium-ion battery cathode materials to manufacturing feedstocks. Nat. Chem. Eng. 2, 142\u2013151 (2025).","journal-title":"Nat. Chem. Eng."},{"key":"1298_CR3","doi-asserted-by":"crossref","unstructured":"Kang, G. et al. Electromagnetic interference shielding using metal and MXene thin films. Nature 647, 356\u2013363 (2025).","DOI":"10.1038\/s41586-025-09699-0"},{"key":"1298_CR4","doi-asserted-by":"publisher","first-page":"eabn3103","DOI":"10.1126\/science.abn3103","volume":"376","author":"Y Yao","year":"2022","unstructured":"Yao, Y. et al. High-entropy nanoparticles: synthesis\u2013structure\u2013property relationships and data-driven discovery. Science 376, eabn3103 (2022).","journal-title":"Science"},{"key":"1298_CR5","doi-asserted-by":"publisher","first-page":"331","DOI":"10.1038\/s41578-019-0101-8","volume":"4","author":"AR Oganov","year":"2019","unstructured":"Oganov, A. R., Pickard, C. J., Zhu, Q. & Needs, R. J. Structure prediction drives materials discovery. Nat. Rev. Mater. 4, 331\u2013348 (2019).","journal-title":"Nat. Rev. Mater."},{"key":"1298_CR6","doi-asserted-by":"publisher","first-page":"514","DOI":"10.1038\/s44160-022-00096-3","volume":"1","author":"AK Cheetham","year":"2022","unstructured":"Cheetham, A. K., Seshadri, R. & Wudl, F. Chemical synthesis and materials discovery. Nat. Synth. 1, 514\u2013520 (2022).","journal-title":"Nat. Synth."},{"key":"1298_CR7","doi-asserted-by":"publisher","first-page":"86","DOI":"10.1038\/s41586-023-06734-w","volume":"624","author":"NJ Szymanski","year":"2023","unstructured":"Szymanski, N. J. et al. An autonomous laboratory for the accelerated synthesis of novel materials. Nature 624, 86\u201391 (2023).","journal-title":"Nature"},{"key":"1298_CR8","doi-asserted-by":"publisher","first-page":"80","DOI":"10.1038\/s41586-023-06735-9","volume":"624","author":"A Merchant","year":"2023","unstructured":"Merchant, A. et al. Scaling deep learning for materials discovery. Nature 624, 80\u201385 (2023).","journal-title":"Nature"},{"key":"1298_CR9","doi-asserted-by":"publisher","first-page":"78","DOI":"10.1126\/science.abo4940","volume":"378","author":"Z Rao","year":"2022","unstructured":"Rao, Z. et al. Machine learning-enabled high-entropy alloy discovery. Science 378, 78\u201385 (2022).","journal-title":"Science"},{"key":"1298_CR10","doi-asserted-by":"publisher","first-page":"718","DOI":"10.1038\/s43588-022-00349-3","volume":"2","author":"C Chen","year":"2022","unstructured":"Chen, C. & Ong, S. P. A universal graph deep learning interatomic potential for the periodic table. Nat. Comput. Sci. 2, 718\u2013728 (2022).","journal-title":"Nat. Comput. Sci."},{"key":"1298_CR11","doi-asserted-by":"publisher","first-page":"1031","DOI":"10.1038\/s42256-023-00716-3","volume":"5","author":"B Deng","year":"2023","unstructured":"Deng, B. et al. CHGNet as a pretrained universal neural network potential for charge-informed atomistic modelling. Nat. Mach. Intell. 5, 1031\u20131041 (2023).","journal-title":"Nat. Mach. Intell."},{"key":"1298_CR12","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/npjcompumats.2015.10","volume":"1","author":"S Kirklin","year":"2015","unstructured":"Kirklin, S. et al. The Open Quantum Materials Database (OQMD): assessing the accuracy of DFT formation energies. npj Comput. Mater. 1, 1\u201315 (2015).","journal-title":"npj Comput. Mater."},{"key":"1298_CR13","doi-asserted-by":"publisher","first-page":"209","DOI":"10.1016\/j.commatsci.2014.10.037","volume":"97","author":"S Ong","year":"2015","unstructured":"Ong, S. et al. The Materials Project: a materials genome approach to accelerating materials innovation. Comput. Mater. Sci. 97, 209\u2013215 (2015).","journal-title":"Comput. Mater. Sci."},{"key":"1298_CR14","doi-asserted-by":"publisher","first-page":"707","DOI":"10.1038\/s41586-023-05773-7","volume":"616","author":"KJ Kanarik","year":"2023","unstructured":"Kanarik, K. J. et al. Human-machine collaboration for improving semiconductor process development. Nature 616, 707\u2013711 (2023).","journal-title":"Nature"},{"key":"1298_CR15","doi-asserted-by":"publisher","DOI":"10.1038\/s41524-025-01568-8","volume":"11","author":"X Tan","year":"2025","unstructured":"Tan, X. et al. Machine learning and high-throughput computational guided development of high temperature oxidation-resisting Ni\u2013Co\u2013Cr\u2013Al\u2013Fe based high-entropy alloys. npj Comput. Mater. 11, 93 (2025).","journal-title":"npj Comput. Mater."},{"key":"1298_CR16","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-47997-9","volume":"15","author":"M Steiner","year":"2024","unstructured":"Steiner, M. & Reiher, M. A human-machine interface for automatic exploration of chemical reaction networks. Nat. Commun. 15, 3680 (2024).","journal-title":"Nat. Commun."},{"key":"1298_CR17","doi-asserted-by":"publisher","first-page":"570","DOI":"10.1038\/s41586-023-06792-0","volume":"624","author":"DA Boiko","year":"2023","unstructured":"Boiko, D. A., MacKnight, R., Kline, B. & Gomes, G. Autonomous chemical research with large language models. Nature 624, 570\u2013578 (2023).","journal-title":"Nature"},{"key":"1298_CR18","doi-asserted-by":"publisher","first-page":"2502771","DOI":"10.1002\/adma.202502771","volume":"37","author":"W Yuan","year":"2025","unstructured":"Yuan, W., Chen, G., Wang, Z. & You, F. Empowering generalist material intelligence with large language models. Adv. Mater. 37, 2502771 (2025).","journal-title":"Adv. Mater."},{"key":"1298_CR19","doi-asserted-by":"publisher","first-page":"437","DOI":"10.1038\/s42256-025-00994-z","volume":"7","author":"Y Zheng","year":"2025","unstructured":"Zheng, Y. et al. Large language models for scientific discovery in molecular property prediction. Nat. Mach. Intell. 7, 437\u2013447 (2025).","journal-title":"Nat. Mach. Intell."},{"key":"1298_CR20","doi-asserted-by":"publisher","first-page":"2514","DOI":"10.1039\/D4SC03921A","volume":"16","author":"MC Ramos","year":"2025","unstructured":"Ramos, M. C., Collison, C. J. & White, A. D. A review of large language models and autonomous agents in chemistry. Chem. Science 16, 2514\u20132572 (2025).","journal-title":"Chem. Science"},{"key":"1298_CR21","doi-asserted-by":"publisher","first-page":"525","DOI":"10.1038\/s42256-024-00832-8","volume":"6","author":"A M. Bran","year":"2024","unstructured":"M. Bran, A. et al. Augmenting large language models with chemistry tools. Nat. Mach. Intell. 6, 525\u2013535 (2024).","journal-title":"Nat. Mach. Intell."},{"key":"1298_CR22","doi-asserted-by":"publisher","first-page":"952","DOI":"10.1038\/s43588-025-00836-3","volume":"5","author":"N Alampara","year":"2025","unstructured":"Alampara, N. et al. Probing the limitations of multimodal language models for chemistry and materials research. Nat. Comput. Sci. 5, 952\u2013961 (2025).","journal-title":"Nat. Comput. Sci."},{"key":"1298_CR23","doi-asserted-by":"publisher","first-page":"991","DOI":"10.1038\/s42256-025-01058-y","volume":"7","author":"S Miret","year":"2025","unstructured":"Miret, S. & Krishnan, N. A. Enabling large language models for real-world materials discovery. Nat. Mach. Intell. 7, 991\u2013998 (2025).","journal-title":"Nat. Mach. Intell."},{"key":"1298_CR24","doi-asserted-by":"publisher","first-page":"277","DOI":"10.1038\/s42254-023-00581-4","volume":"5","author":"A Birhane","year":"2023","unstructured":"Birhane, A., Kasirzadeh, A., Leslie, D. & Wachter, S. Science in the age of large language models. Nat. Rev. Phys. 5, 277\u2013280 (2023).","journal-title":"Nat. Rev. Phys."},{"key":"1298_CR25","doi-asserted-by":"publisher","first-page":"1257","DOI":"10.1039\/D4DD00074A","volume":"3","author":"G Lei","year":"2024","unstructured":"Lei, G., Docherty, R. & Cooper, S. J. Materials science in the era of large language models: a perspective. Digit. Discov. 3, 1257\u20131272 (2024).","journal-title":"Digit. Discov."},{"key":"1298_CR26","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-54639-7","volume":"15","author":"LM Antunes","year":"2024","unstructured":"Antunes, L. M., Butler, K. T. & Grau-Crespo, R. Crystal structure generation with autoregressive large language modeling. Nat. Commun. 15, 10570 (2024).","journal-title":"Nat. Commun."},{"key":"1298_CR27","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-61778-y","volume":"16","author":"Z Song","year":"2025","unstructured":"Song, Z., Lu, S., Ju, M., Zhou, Q. & Wang, J. Accurate prediction of synthesizability and precursors of 3D crystal structures via large language models. Nat. Commun. 16, 6530 (2025).","journal-title":"Nat. Commun."},{"key":"1298_CR28","doi-asserted-by":"publisher","first-page":"39113","DOI":"10.1021\/jacs.5c07743","volume":"147","author":"J Choi","year":"2025","unstructured":"Choi, J., Kim, S. & Jung, Y. Synthesis-aware materials redesign via large language models. J. Am. Chem. Soc. 147, 39113\u201339122 (2025).","journal-title":"J. Am. Chem. Soc."},{"key":"1298_CR29","doi-asserted-by":"publisher","first-page":"50117","DOI":"10.52202\/075280-2180","volume":"36","author":"Y Zhuang","year":"2023","unstructured":"Zhuang, Y., Yu, Y., Wang, K., Sun, H. & Zhang, C. ToolQA: a dataset for LLM question answering with external tools. Adv. Neural Inf. Process. Syst. 36, 50117\u201350143 (2023).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"1298_CR30","doi-asserted-by":"crossref","unstructured":"Lu, J. et al. ToolSandbox: a stateful, conversational, interactive evaluation benchmark for LLM tool use capabilities. In Findings of the Association for Computational Linguistics: NAACL 2025 (eds Chiruzzo, L., Ritter, A. & Wang, L.) 1160\u20131183 (ACL, 2025).","DOI":"10.18653\/v1\/2025.findings-naacl.65"},{"key":"1298_CR31","doi-asserted-by":"publisher","first-page":"198343","DOI":"10.1007\/s11704-024-40678-2","volume":"19","author":"C Qu","year":"2025","unstructured":"Qu, C. et al. Tool learning with large language models: a survey. Front. Comput. Sci. 19, 198343 (2025).","journal-title":"Front. Comput. Sci."},{"key":"1298_CR32","doi-asserted-by":"crossref","unstructured":"Xiao, C. et al. Densing law of LLMs. Nat. Mach. Intell. 7, 1823\u20131833 (2025).","DOI":"10.1038\/s42256-025-01137-0"},{"key":"1298_CR33","unstructured":"Barros, T. D. S., Giroire, F., Aparicio-Pardo, R. & Moulierac, J. Small is sufficient: reducing the world AI energy consumption through model selection. Preprint at https:\/\/arxiv.org\/abs\/2510.01889 (2025)."},{"key":"1298_CR34","unstructured":"Li, K. et al. WebSailor-V2: bridging the chasm to proprietary agents via synthetic data and scalable reinforcement learning. In The 14th International Conference on Learning Representations (ICLR, 2026)."},{"key":"1298_CR35","unstructured":"Dong, G. et al. Agentic reinforced policy optimization. In The Fourteenth International Conference on Learning Representations (ICLR, 2026)."},{"key":"1298_CR36","unstructured":"Shang, N. et al. rStar2-agent: agentic reasoning technical report. Preprint at https:\/\/arxiv.org\/abs\/2508.20722 (2025)."},{"key":"1298_CR37","first-page":"113222","volume":"38","author":"Q Yu","year":"2026","unstructured":"Yu, Q. et al. DAPO: an open-source LLM reinforcement learning system at scale. Adv. Neural Inf. Process. Syst. 38, 113222\u2013113244 (2026).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"1298_CR38","doi-asserted-by":"publisher","first-page":"eadg8180","DOI":"10.1126\/sciadv.adg8180","volume":"9","author":"T He","year":"2023","unstructured":"He, T. et al. Precursor recommendation for inorganic synthesis by machine learning materials similarity from scientific literature. Sci. Adv. 9, eadg8180 (2023).","journal-title":"Sci. Adv."},{"key":"1298_CR39","doi-asserted-by":"publisher","first-page":"107","DOI":"10.1038\/s41560-021-00941-3","volume":"7","author":"TJ Jacobsson","year":"2022","unstructured":"Jacobsson, T. J. et al. An open-access database and analysis tool for perovskite solar cells based on the FAIR data principles. Nat. Energy 7, 107\u2013115 (2022).","journal-title":"Nat. Energy"},{"key":"1298_CR40","doi-asserted-by":"publisher","DOI":"10.1038\/s41524-025-01538-0","volume":"11","author":"EO Pyzer-Knapp","year":"2025","unstructured":"Pyzer-Knapp, E. O. et al. Foundation models for materials discovery-current state and future directions. npj Comput. Mater. 11, 61 (2025).","journal-title":"npj Comput. Mater."},{"key":"1298_CR41","unstructured":"Rein, D. et al. GPQA: a graduate-level Google-proof Q&A benchmark. In First Conference on Language Modeling (ICLM, 2024)."},{"key":"1298_CR42","doi-asserted-by":"crossref","unstructured":"Center for AI Safety, Scale AI & HLE Contributors Consortium. A benchmark of expert-level academic questions to assess AI capabilities. Nature 649, 1139\u20131146 (2026).","DOI":"10.1038\/s41586-025-09962-4"},{"key":"1298_CR43","unstructured":"Liu, A. et al. DeepSeek-v3 technical report. Preprint at https:\/\/arxiv.org\/abs\/2412.19437 (2024)."},{"key":"1298_CR44","doi-asserted-by":"publisher","first-page":"1010","DOI":"10.1038\/s42256-025-01066-y","volume":"7","author":"Y Zhang","year":"2025","unstructured":"Zhang, Y. et al. Large language models to accelerate organic chemistry synthesis. Nat. Mach. Intell. 7, 1010\u20131022 (2025).","journal-title":"Nat. Mach. Intell."},{"key":"1298_CR45","doi-asserted-by":"publisher","unstructured":"Hou, X., Zhao, Y., Wang, S. & Wang, H. Model context protocol (MCP): landscape, security threats, and future research directions. ACM Trans. Softw. Eng. Methodol. https:\/\/doi.org\/10.1145\/3796519 (2026).","DOI":"10.1145\/3796519"},{"key":"1298_CR46","unstructured":"Li, K. et al. WebSailor: navigating super-human reasoning for web agent. Preprint at https:\/\/arxiv.org\/abs\/2507.02592 (2025)."},{"key":"1298_CR47","first-page":"120091","volume":"38","author":"X Li","year":"2026","unstructured":"Li, X. et al. WebThinker: empowering large reasoning models with deep research capability. Adv. Neural Inf. Process. Syst. 38, 120091\u2013120131 (2026).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"1298_CR48","unstructured":"Bai, L. et al. Intern-s1: s scientific multimodal foundation model. Preprint at https:\/\/arxiv.org\/abs\/2508.15763 (2025)."},{"key":"1298_CR49","doi-asserted-by":"publisher","first-page":"962","DOI":"10.1038\/s43588-025-00849-y","volume":"5","author":"K Ding","year":"2025","unstructured":"Ding, K. et al. SciToolAgent: a knowledge-graph-driven scientific agent for multitool integration. Nat. Comput. Sci. 5, 962\u2013972 (2025).","journal-title":"Nat. Comput. Sci."},{"key":"1298_CR50","doi-asserted-by":"crossref","unstructured":"Qian, C. et al. ToolRL: reward is all tool learning needs. In Neural Information Processing Systems (NeurIPS, 2025).","DOI":"10.52202\/085713-3524"},{"key":"1298_CR51","unstructured":"Cui, G. et al. Process reinforcement through implicit rewards. In Transactions on Machine Learning Research (TMLR, 2026)."},{"key":"1298_CR52","doi-asserted-by":"crossref","unstructured":"Skalse, J., Howe, N., Krasheninnikov, D. & Krueger, D. Defining and characterizing reward gaming. In Neural Information Processing Systems (NeurIPS, 2022).","DOI":"10.52202\/068431-0687"},{"key":"1298_CR53","doi-asserted-by":"crossref","unstructured":"Guo, D. et al. DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning. Nature 645, 633\u2013638 (2025).","DOI":"10.1038\/s41586-025-09422-z"},{"key":"1298_CR54","unstructured":"Kuhn, L., Gal, Y. & Farquhar, S. Semantic uncertainty: linguistic invariances for uncertainty estimation in natural language generation. In International Conference on Learning Representations (ICLR, 2023)."},{"key":"1298_CR55","doi-asserted-by":"crossref","unstructured":"Ouyang, L. et al. Training language models to follow instructions with human feedback. In Neural Information Processing Systems (NeurIPS, 2022).","DOI":"10.52202\/068431-2011"},{"key":"1298_CR56","doi-asserted-by":"publisher","first-page":"6709","DOI":"10.1021\/acs.inorgchem.9b03421","volume":"59","author":"Y Nakamura","year":"2020","unstructured":"Nakamura, Y. et al. Crystal systems and lattice Parameters of $${\\mathrm{CH}}_{3}{\\mathrm{NH}}_{3}\\mathrm{Pb}{({{\\rm{I}}}_{1-x}{\\mathrm{Br}}_{x})}_{3}$$ determined using single crystals: validity of Vegard\u2019s law. Inorg. Chem. 59, 6709\u20136716 (2020).","journal-title":"Inorg. Chem."},{"key":"1298_CR57","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-021-21139-x","volume":"12","author":"G Ermolaev","year":"2021","unstructured":"Ermolaev, G. et al. Giant optical anisotropy in transition metal dichalcogenides for next-generation photonics. Nat. Commun. 12, 854 (2021).","journal-title":"Nat. Commun."},{"key":"1298_CR58","unstructured":"Wang, B. et al. MinerU: an open-source solution for precise document content extraction. Preprint at https:\/\/arxiv.org\/abs\/2409.18839 (2024)."},{"key":"1298_CR59","unstructured":"Goel, A. LangExtract: a library for extracting structured data from language models. GitHub https:\/\/github.com\/google\/langextract (2025)."},{"key":"1298_CR60","doi-asserted-by":"crossref","unstructured":"Zhao, Y. et al. SWIFT: a scalable lightweight infrastructure for fine-tuning. In Proc. AAAI Conference on Artificial Intelligence (eds Walsh, T., Shah, J. & Kolter, Z.) 29733\u201329735 (AAAI, 2025).","DOI":"10.1609\/aaai.v39i28.35383"},{"key":"1298_CR61","unstructured":"Dao, T. FlashAttention-2: faster attention with better parallelism and work partitioning. In The Twelfth International Conference on Learning Representations (eds Kim, B. et al.) (ICLR, 2024)."},{"key":"1298_CR62","unstructured":"Loshchilov, I. & Hutter, F. Decoupled weight decay regularization. In 7th International Conference on Learning Representations (ICLR, 2019)."},{"key":"1298_CR63","doi-asserted-by":"crossref","unstructured":"Sheng, G. et al. HybridFlow: a flexible and efficient RLHF framework. In Proc. 20th European Conference on Computer Systems (eds Lawall, J. et al.) 1279\u20131297 (ACM, 2025).","DOI":"10.1145\/3689031.3696075"},{"key":"1298_CR64","doi-asserted-by":"crossref","unstructured":"Kwon, W. et al. Efficient memory management for large language model serving with PagedAttention. In Proc. 29th Symposium on Operating Systems Principles (eds Flinn, J. et al.) 611\u2013626 (ACM, 2023).","DOI":"10.1145\/3600006.3613165"},{"key":"1298_CR65","doi-asserted-by":"publisher","unstructured":"Shi, T. et al. Mat-252K-SFT. HuggingFace https:\/\/doi.org\/10.57967\/hf\/9652 (2026).","DOI":"10.57967\/hf\/9652"},{"key":"1298_CR66","doi-asserted-by":"publisher","unstructured":"Shi, T. et al. Mat-20K-RL. HuggingFace https:\/\/doi.org\/10.57967\/hf\/9653 (2026).","DOI":"10.57967\/hf\/9653"},{"key":"1298_CR67","doi-asserted-by":"publisher","unstructured":"Shi, T. et al. MatBrain source code. Zenodo https:\/\/doi.org\/10.5281\/zenodo.21507186 (2026).","DOI":"10.5281\/zenodo.21507186"}],"container-title":["Nature Machine Intelligence"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s42256-026-01298-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s42256-026-01298-6","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s42256-026-01298-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T22:02:42Z","timestamp":1789682562000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s42256-026-01298-6"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,10]]},"references-count":67,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2026,9]]}},"alternative-id":["1298"],"URL":"https:\/\/doi.org\/10.1038\/s42256-026-01298-6","relation":{},"ISSN":["2522-5839"],"issn-type":[{"value":"2522-5839","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,10]]},"assertion":[{"value":"12 February 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"30 July 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"10 September 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Ethics"}}]}}