{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,7]],"date-time":"2026-03-07T17:44:17Z","timestamp":1772905457629,"version":"3.50.1"},"reference-count":84,"publisher":"Walter de Gruyter GmbH","issue":"3","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,3,26]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Model compression is key for deploying Deep Neural Networks on resource-constrained hardware. Its application to Neural Network Controllers (NNCs) is challenging because it can compromise control-theoretic properties and performance, a critical issue for modern controllers that use latent-space models. This paper surveys and empirically evaluates compression techniques for these models, using an MNIST autoencoder and a Temporal Difference Model Predictive Control agent as test cases across diverse hardware. We find that general compression techniques apply to latent-space models and that careful compression can preserve the theoretical properties of NNCs. Specific findings indicate that quantization can increase latency on non-specialized hardware, fine-tuning is crucial for performance recovery, and hybrid methods yield the best trade-offs.<\/jats:p>","DOI":"10.1515\/auto-2025-0107","type":"journal-article","created":{"date-parts":[[2026,3,7]],"date-time":"2026-03-07T01:43:52Z","timestamp":1772847832000},"page":"212-232","source":"Crossref","is-referenced-by-count":0,"title":["Model compression for\u00a0neural network controllers: a tutorial survey with\u00a0a focus on\u00a0controllers with\u00a0latent state space"],"prefix":"10.1515","volume":"74","author":[{"given":"Ganesh","family":"Sundaram","sequence":"first","affiliation":[{"name":"Institute of Electromobility , 26562 RPTU University Kaiserslautern-Landau , 67663 Kaiserslautern , Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jonas","family":"Ulmen","sequence":"additional","affiliation":[{"name":"Institute of Electromobility , 26562 RPTU University Kaiserslautern-Landau , 67663 Kaiserslautern , Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Daniel","family":"G\u00f6rges","sequence":"additional","affiliation":[{"name":"Institute of Electromobility , 26562 RPTU University Kaiserslautern-Landau , 67663 Kaiserslautern , Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"374","published-online":{"date-parts":[[2026,3,9]]},"reference":[{"key":"2026030701434641988_j_auto-2025-0107_ref_001","unstructured":"Y. Cheng, D. Wang, P. Zhou, and T. Zhang, \u201cA survey of model compression and acceleration for deep neural networks,\u201d arXiv:1710.09282, 2017."},{"key":"2026030701434641988_j_auto-2025-0107_ref_002","unstructured":"S. Han, J. Pool, J. Tran, and W. Dally, \u201cLearning both weights and connections for efficient neural network,\u201d Adv. Neural Inf. Process. Syst., vol.\u00a028, pp.\u00a01135\u20131143, 2015."},{"key":"2026030701434641988_j_auto-2025-0107_ref_003","unstructured":"W. Wen, C. Wu, Y. Wang, Y. Chen, and H. Li, \u201cLearning structured sparsity in deep neural networks,\u201d Adv. Neural Inf. Process. Syst., vol.\u00a029, pp.\u00a02074\u20132082, 2016."},{"key":"2026030701434641988_j_auto-2025-0107_ref_004","doi-asserted-by":"crossref","unstructured":"Z. Liu, J. Li, Z. Shen, G. Huang, S. Yan, and C. Zhang, \u201cLearning efficient convolutional networks through network slimming,\u201d in Proceedings of the 15th IEEE International Conference on Computer Vision, Venice, Italy (Palazzo del Cinema), IEEE Computer Society, 2017, pp.\u00a02736\u20132744.","DOI":"10.1109\/ICCV.2017.298"},{"key":"2026030701434641988_j_auto-2025-0107_ref_005","doi-asserted-by":"crossref","unstructured":"Y. He, X. Zhang, and J. Sun, \u201cChannel pruning for accelerating very deep neural networks,\u201d in Proceedings of the 15th IEEE International Conference on Computer Vision, Venice, Italy (Palazzo del Cinema), IEEE Computer Society, 2017, pp.\u00a01389\u20131397.","DOI":"10.1109\/ICCV.2017.155"},{"key":"2026030701434641988_j_auto-2025-0107_ref_006","unstructured":"D. Molchanov, A. Ashukha, and D. Vetrov, \u201cVariational dropout sparsifies deep neural networks,\u201d in Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia, Proceedings of Machine Learning Research (PMLR) \/ JMLR, 2017, pp.\u00a02498\u20132507."},{"key":"2026030701434641988_j_auto-2025-0107_ref_007","unstructured":"J. Frankle and M. Carbin, \u201cThe lottery ticket hypothesis: Finding sparse, trainable neural networks,\u201d in Proceeding of the 7th International Conference on Learning Representations, Louisiana, USA, OpenReview.net and Curran Associates, Inc., 2019."},{"key":"2026030701434641988_j_auto-2025-0107_ref_008","doi-asserted-by":"crossref","unstructured":"D. Mocanu, E. Mocanu, P. Stone, P. Nguyen, M. Gibescu, and A. Liotta, \u201cScalable training of artificial neural networks with adaptive sparse connectivity inspired by network science,\u201d Nat. Commun., vol.\u00a09, 2018, https:\/\/doi.org\/10.1038\/s41467-018-04316-3.","DOI":"10.1038\/s41467-018-04316-3"},{"key":"2026030701434641988_j_auto-2025-0107_ref_009","doi-asserted-by":"crossref","unstructured":"G. Sundaram, J. Ulmen, and D. G\u00f6rges, \u201cEnhanced pruning strategy for multi-component neural architectures using component-aware graph analysis,\u201d in Proceedings of the 1st IFAC Joint Conference on Computers, Cognition and Communication, Padova, Italy, Elsevier, 2025.","DOI":"10.1016\/j.ifacol.2025.11.934"},{"key":"2026030701434641988_j_auto-2025-0107_ref_010","doi-asserted-by":"crossref","unstructured":"B. Jacob et al.., \u201cQuantization and training of neural networks for efficient integer-arithmetic-only inference,\u201d in Proceedings of the 31st IEEE Conference on Computer Vision and Pattern Recognition, Utah, USA, IEEE Computer Society, 2018, pp.\u00a02704\u20132713.","DOI":"10.1109\/CVPR.2018.00286"},{"key":"2026030701434641988_j_auto-2025-0107_ref_011","unstructured":"R. Krishnamoorthi, \u201cQuantizing deep convolutional networks for efficient inference: A whitepaper,\u201d arXiv:1806.08342, 2018."},{"key":"2026030701434641988_j_auto-2025-0107_ref_012","unstructured":"M. Courbariaux, I. Hubara, D. Soudry, R. El-Yaniv, and Y. Bengio, \u201cBinarized neural networks: Training deep neural networks with weights and activations constrained to+ 1 or-1,\u201d arXiv:1602.02830, 2016."},{"key":"2026030701434641988_j_auto-2025-0107_ref_013","unstructured":"K. Wang, Z. Liu, Y. Lin, J. Lin, and S. Han, \u201cAMC: AutoML for model compression and acceleration on mobile devices,\u201d in Proceedings of the 15th European Conference on Computer Vision, Munich, Germany, Springer International, 2019, pp.\u00a0784\u2013800."},{"key":"2026030701434641988_j_auto-2025-0107_ref_014","unstructured":"S. K. Esser, J. L. McKinstry, D. Bablani, R. Appuswamy, and D. S. Modha, \u201cLearned step size quantization,\u201d in Proceedings of the 8th International Conference on Learning Representations, Addis Ababa, Ethiopia, OpenReview.net, 2020."},{"key":"2026030701434641988_j_auto-2025-0107_ref_015","doi-asserted-by":"crossref","unstructured":"M. Rastegari, V. Ordonez, J. Redmon, and A. Farhadi, \u201cXNOR-Net: Imagenet classification using binary convolutional neural networks,\u201d in Proceedings of the 14th European Conference on Computer Vision, Amsterdam, The Netherlands, Springer International, 2016, pp.\u00a0525\u2013542.","DOI":"10.1007\/978-3-319-46493-0_32"},{"key":"2026030701434641988_j_auto-2025-0107_ref_016","unstructured":"G. Hinton, O. Vinyals, and J. Dean, \u201cDistilling the knowledge in a neural network,\u201d arXiv:1503.02531, 2015."},{"key":"2026030701434641988_j_auto-2025-0107_ref_017","unstructured":"A. Romero, N. Ballas, S. E. Kahou, A. Chassang, C. Gatta, and Y. Bengio, \u201cFitNets: Hints for thin deep nets,\u201d in Proceedings of the 3rd International Conference on Learning Representations, San Diego, California, USA, Computational and Biological Learning Society (CBLS), 2015."},{"key":"2026030701434641988_j_auto-2025-0107_ref_018","unstructured":"S. Zagoruyko and N. Komodakis, \u201cPaying more attention to attention: Improving the performance of convolutional neural networks via attention transfer,\u201d in Proceedings of the 15th International Conference on Learning Representations, Toulon, France, OpenReview.net, 2017."},{"key":"2026030701434641988_j_auto-2025-0107_ref_019","doi-asserted-by":"crossref","unstructured":"Y. Zhang, T. Xiang, T. M. Hospedales, and H. Lu, \u201cDeep mutual learning,\u201d in Proceedings of the 31st IEEE Conference on Computer Vision and Pattern Recognition, Utah, USA, The IEEE Computer Society and the Computer Vision Foundation, 2018, pp.\u00a04320\u20134328.","DOI":"10.1109\/CVPR.2018.00454"},{"key":"2026030701434641988_j_auto-2025-0107_ref_020","unstructured":"T. Furlanello, Z. Lipton, M. Tschannen, L. Itti, and A. Anandkumar, \u201cBorn again neural networks,\u201d in Proceedings of the 35th International Conference on Machine Learning, Stockholm, Sweden, Proceedings of Machine Learning Research (PMLR), 2018, pp.\u00a01607\u20131616."},{"key":"2026030701434641988_j_auto-2025-0107_ref_021","doi-asserted-by":"crossref","unstructured":"D. Chen, J.-P. Mei, C. Wang, Y. Feng, and C. Chen, \u201cImproved knowledge distillation via teacher assistant,\u201d in Proceedings of the 34th AAAI Conference on Artificial Intelligence, vol.\u00a034, New York, USA, Association for the Advancement of Artificial Intelligence (AAAI), 2020, pp.\u00a05191\u20135198.","DOI":"10.1609\/aaai.v34i04.5963"},{"key":"2026030701434641988_j_auto-2025-0107_ref_022","unstructured":"E. L. Denton, W. Zaremba, J. Bruna, Y. LeCun, and R. Fergus, \u201cExploiting linear structure within convolutional networks for efficient evaluation,\u201d Adv. Neural Inf. Process. Syst., vol.\u00a027, pp.\u00a01269\u20131277, 2014."},{"key":"2026030701434641988_j_auto-2025-0107_ref_023","doi-asserted-by":"crossref","unstructured":"M. Jaderberg, A. Vedaldi, and A. Zisserman, \u201cSpeeding up convolutional neural networks with low rank expansions,\u201d in British Machine Vision Conference, Nottingham, UK, BMVA Press, 2014.","DOI":"10.5244\/C.28.88"},{"key":"2026030701434641988_j_auto-2025-0107_ref_024","unstructured":"V. Lebedev, Y. Ganin, M. Rakhuba, I. Oseledets, and V. Lempitsky, \u201cSpeeding-up convolutional neural networks using fine-tuned CP-decomposition,\u201d in Proceedings of the 3rd International Conference on Learning Representations, San Diego, California, USA, International Conference on Learning Representations (ICLR), 2015."},{"key":"2026030701434641988_j_auto-2025-0107_ref_025","unstructured":"Y.-D. Kim, E. Park, S. Yoo, T. Choi, L. Yang, and D. Shin, \u201cCompression of deep convolutional neural networks for fast and low power mobile applications,\u201d in Proceedings of the 15th International Conference on Learning Representations, San Juan, Puerto Rico, OpenReview, 2016."},{"key":"2026030701434641988_j_auto-2025-0107_ref_026","doi-asserted-by":"crossref","unstructured":"X. Yu, T. Liu, X. Wang, and D. Tao, \u201cOn compressing deep models by low rank and sparse decomposition,\u201d in Proceedings of the 30th IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, Hawaii, IEEE Computer Society, 2017, pp.\u00a07370\u20137379.","DOI":"10.1109\/CVPR.2017.15"},{"key":"2026030701434641988_j_auto-2025-0107_ref_027","doi-asserted-by":"crossref","unstructured":"A. C. Antoulas, Approximation of Large-Scale Dynamical Systems, Philadelphia, PA, SIAM, 2005.","DOI":"10.1137\/1.9780898718713"},{"key":"2026030701434641988_j_auto-2025-0107_ref_028","doi-asserted-by":"crossref","unstructured":"P. Benner, S. Gugercin, and K. Willcox, \u201cA survey of projection-based model reduction methods for parametric dynamical systems,\u201d SIAM Review, vol.\u00a057, no.\u00a04, pp.\u00a0483\u2013531, 2015, https:\/\/doi.org\/10.1137\/130932715.","DOI":"10.1137\/130932715"},{"key":"2026030701434641988_j_auto-2025-0107_ref_029","doi-asserted-by":"crossref","unstructured":"A. Quarteroni, A. Manzoni, and F. Negri, Reduced Basis Methods for Partial Differential Equations: An Introduction, Berlin, Springer, 2016.","DOI":"10.1007\/978-3-319-15431-2"},{"key":"2026030701434641988_j_auto-2025-0107_ref_030","doi-asserted-by":"crossref","unstructured":"E. J. Parish and K. Duraisamy, \u201cA paradigm for data-driven predictive modeling using field inversion and machine learning,\u201d J. Comput. Phys., vol.\u00a0305, pp.\u00a0758\u2013774, 2016, https:\/\/doi.org\/10.1016\/j.jcp.2015.11.012.","DOI":"10.1016\/j.jcp.2015.11.012"},{"key":"2026030701434641988_j_auto-2025-0107_ref_031","doi-asserted-by":"crossref","unstructured":"B. Recht, \u201cA tour of reinforcement learning: The view from continuous control,\u201d Ann. Rev. Control, Rob., Autonom. Syst., vol.\u00a02, pp.\u00a0253\u2013279, 2019, https:\/\/doi.org\/10.1146\/annurev-control-053018-023825.","DOI":"10.1146\/annurev-control-053018-023825"},{"key":"2026030701434641988_j_auto-2025-0107_ref_032","unstructured":"M. Fazlyab, A. Robey, H. Hassani, M. Morari, and G. Pappas, \u201cEfficient and accurate estimation of Lipschitz constants for deep neural networks,\u201d Adv. Neural Inf. Process. Syst., vol.\u00a032, 2019."},{"key":"2026030701434641988_j_auto-2025-0107_ref_033","doi-asserted-by":"crossref","unstructured":"M. Revay, R. Wang, and I. R. Manchester, \u201cRecurrent equilibrium networks: Flexible dynamic models with guaranteed stability and robustness,\u201d IEEE Trans. Autom. Control, vol.\u00a069, no.\u00a05, pp.\u00a02855\u20132870, 2023, https:\/\/doi.org\/10.1109\/tac.2023.3294101.","DOI":"10.1109\/TAC.2023.3294101"},{"key":"2026030701434641988_j_auto-2025-0107_ref_034","unstructured":"S. Tu and B. Recht, \u201cThe gap between model-based and model-free methods on the linear quadratic regulator: An asymptotic viewpoint,\u201d in Proceedings of the 32nd Conference on Learning Theory, Arizona, USA, Proceedings of Machine Learning Research (PMLR), 2019, pp.\u00a03036\u20133083."},{"key":"2026030701434641988_j_auto-2025-0107_ref_035","doi-asserted-by":"crossref","unstructured":"B. Karg and S. Lucia, \u201cEfficient representation and approximation of model predictive control laws via deep learning,\u201d IEEE Trans. Cybern., vol.\u00a050, no.\u00a09, pp.\u00a03866\u20133878, 2020, https:\/\/doi.org\/10.1109\/tcyb.2020.2999556.","DOI":"10.1109\/TCYB.2020.2999556"},{"key":"2026030701434641988_j_auto-2025-0107_ref_036","unstructured":"A. A. Rusu et al.., \u201cPolicy distillation,\u201d in Proceedings of the 4th International Conference on Learning Representations, San Juan, Puerto Rico, International Conference on Learning Representations, 2016."},{"key":"2026030701434641988_j_auto-2025-0107_ref_037","unstructured":"W. M. Czarnecki, R. Pascanu, S. Osindero, S. Jayakumar, G. Swirszcz, and M. Jaderberg, \u201cDistilling policy distillation,\u201d in Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics, Okinawa, Japan, JMLR (Journal of Machine Learning Research), 2019, pp.\u00a01331\u20131340."},{"key":"2026030701434641988_j_auto-2025-0107_ref_038","unstructured":"M. Tan and Q. Le, \u201cEfficientnet: Rethinking model scaling for convolutional neural networks,\u201d in Proceedings of the 36th International Conference on Machine Learning, California, USA, Proceedings of Machine Learning Research (PMLR), 2019, pp.\u00a06105\u20136114."},{"key":"2026030701434641988_j_auto-2025-0107_ref_039","unstructured":"A. Kumar, J. Fu, G. Tucker, and S. Levine, \u201cData-efficient reinforcement learning with learned controllers,\u201d arXiv:2012.10735, 2020."},{"key":"2026030701434641988_j_auto-2025-0107_ref_040","unstructured":"A. Ashok, N. Rhinehart, F. Beainy, and K. M. Kitani, \u201cN2n learning: Network to network compression via policy gradient reinforcement learning,\u201d in Proceedings of the 6th International Conference on Learning Representations, Vancouver, BC, Canada, International Conference on Learning Representations, 2018."},{"key":"2026030701434641988_j_auto-2025-0107_ref_041","doi-asserted-by":"crossref","unstructured":"J. Qiu et al.., \u201cGoing deeper with embedded fpga platform for convolutional neural network,\u201d in Proceedings of the 24th ACM\/SIGDA International Symposium on Field-Programmable Gate Arrays, Monterey, Association for Computing Machinery, 2016, pp.\u00a026\u201335.","DOI":"10.1145\/2847263.2847265"},{"key":"2026030701434641988_j_auto-2025-0107_ref_042","unstructured":"A. G. Howard et al.., \u201cMobilenets: Efficient convolutional neural networks for mobile vision applications,\u201d arXiv:1704.04861, 2017."},{"key":"2026030701434641988_j_auto-2025-0107_ref_043","unstructured":"S. Han, H. Mao, and W. J. Dally, \u201cDeep compression: Compressing deep neural networks with pruning, trained quantization and Huffman coding,\u201d in Proceedings of the 4th International Conference on Learning Representations, San Juan, OpenReview, 2016."},{"key":"2026030701434641988_j_auto-2025-0107_ref_044","doi-asserted-by":"crossref","unstructured":"G. Sundaram, J. Ulmen, A. Haider, and D. G\u00f6rges, \u201cComponent-aware pruning for accelerated control tasks in latent space models,\u201d arXiv:2508.08144, 2025.","DOI":"10.1109\/SII64115.2026.11404702"},{"key":"2026030701434641988_j_auto-2025-0107_ref_045","doi-asserted-by":"crossref","unstructured":"V. Sze, Y.-H. Chen, T.-J. Yang, and J. S. Emer, \u201cEfficient processing of deep neural networks: A tutorial and survey,\u201d Proc. IEEE, vol.\u00a0105, no.\u00a012, pp.\u00a02295\u20132329, 2017, https:\/\/doi.org\/10.1109\/jproc.2017.2761740.","DOI":"10.1109\/JPROC.2017.2761740"},{"key":"2026030701434641988_j_auto-2025-0107_ref_046","doi-asserted-by":"crossref","unstructured":"R. Wilhelm et al.., \u201cThe worst-case execution-time problem\u2014overview of methods and survey of tools,\u201d ACM Trans. Embed. Comput. Syst., vol.\u00a07, no.\u00a03, pp.\u00a01\u201353, 2008, https:\/\/doi.org\/10.1145\/1347375.1347389.","DOI":"10.1145\/1347375.1347389"},{"key":"2026030701434641988_j_auto-2025-0107_ref_047","unstructured":"Y. Kang et al.., \u201cCo-optimizing performance and memory footprint via integrated cpu\/gpu memory management, an implementation on autonomous driving platform,\u201d in 2016 IEEE Real-Time and Embedded Technology and Applications Symposium, Vienna, Institute of Electrical and Electronics Engineers (IEEE), 2016, pp.\u00a01\u201312."},{"key":"2026030701434641988_j_auto-2025-0107_ref_048","unstructured":"B. Zoph and Q. V. Le, \u201cNeural architecture search with reinforcement learning,\u201d in Proceedings of the 5th International Conference on Learning Representations, Toulon, OpenReview, 2017."},{"key":"2026030701434641988_j_auto-2025-0107_ref_049","unstructured":"J. Mellor, J. Turner, A. Storkey, and E. J. Crowley, \u201cNeural architecture search without training,\u201d in International Conference on Machine Learning, PMLR (Virtual Only Conference), 2021, pp.\u00a07588\u20137598."},{"key":"2026030701434641988_j_auto-2025-0107_ref_050","unstructured":"A. Polino, R. Pascanu, and D. Alistarh, \u201cModel compression via distillation and quantization,\u201d in International Conference on Learning Representations, Vancouver, OpenReview, 2018."},{"key":"2026030701434641988_j_auto-2025-0107_ref_051","doi-asserted-by":"crossref","unstructured":"B. Heo, M. Lee, S. Yun, and J. Y. Choi, \u201cA comprehensive overhaul of feature distillation,\u201d in Proceedings of the 17th IEEE\/CVF International Conference on Computer Vision, Seoul, IEEE Computer Society (CPS) \/ IEEE, 2019, pp.\u00a01921\u20131930.","DOI":"10.1109\/ICCV.2019.00201"},{"key":"2026030701434641988_j_auto-2025-0107_ref_052","unstructured":"G. Sundaram, J. Ulmen, A. Haider, and D. G\u00f6rges, \u201cApplication-specific component-aware structured pruning of deep neural networks via soft coefficient optimization,\u201d arXiv:2507.14882, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_053","doi-asserted-by":"crossref","unstructured":"S. Teerapittayanon, B. McDanel, and H.-T. Kung, \u201cBranchynet: Fast inference via early exiting from deep neural networks,\u201d in 2016 23rd International Conference on Pattern Recognition, Cancun, IEEE, 2016, pp.\u00a02464\u20132469.","DOI":"10.1109\/ICPR.2016.7900006"},{"key":"2026030701434641988_j_auto-2025-0107_ref_054","unstructured":"D. Ha and J. Schmidhuber, \u201cWorld models,\u201d Adv. Neural Inf. Process. Syst., vol.\u00a031, 2018."},{"key":"2026030701434641988_j_auto-2025-0107_ref_055","unstructured":"D. Hafner et al.., \u201cLearning latent dynamics for planning from pixels,\u201d in International Conference on Machine Learning, PMLR (Virtual Only Conference), 2019, pp.\u00a02555\u20132565."},{"key":"2026030701434641988_j_auto-2025-0107_ref_056","unstructured":"D. Hafner, T. Lillicrap, M. Norouzi, and J. Ba, \u201cMastering Atari with discrete world models,\u201d in International Conference on Learning Representations, OpenReview (Virtual Only Conference), 2020."},{"key":"2026030701434641988_j_auto-2025-0107_ref_057","unstructured":"D. Hafner, T. Lillicrap, M. Norouzi, and J. Ba, \u201cMastering atari with discrete world models,\u201d arXiv:2010.02193, 2020."},{"key":"2026030701434641988_j_auto-2025-0107_ref_058","unstructured":"K. Asadi, D. Misra, and M. Littman, \u201cLipschitz continuity in model-based reinforcement learning,\u201d in Proceedings of the 35th International Conference on Machine Learning, Stockholm, PMLR, 2018, pp.\u00a0264\u2013273."},{"key":"2026030701434641988_j_auto-2025-0107_ref_059","unstructured":"N. Hansen, X. Wang, and H. Su, \u201cTemporal difference learning for model predictive control,\u201d arXiv:2203.04955, 2022."},{"key":"2026030701434641988_j_auto-2025-0107_ref_060","unstructured":"N. Hansen, H. Su, and X. Wang, \u201cTD-MPC2: Scalable, robust world models for continuous control,\u201d arXiv:2310.16828, 2024."},{"key":"2026030701434641988_j_auto-2025-0107_ref_061","doi-asserted-by":"crossref","unstructured":"C.-H. Ji, H.-B. Choi, J.-S. Heo, J.-B. Kim, H.-K. Lim, and Y.-H. Han, \u201cDual policy-based td-learning for model predictive control,\u201d in 2023 International Conference on Artificial Intelligence in Information and Communication, Kuta, Springer Singapore, 2023, pp.\u00a0254\u2013258.","DOI":"10.1109\/ICAIIC57133.2023.10067037"},{"key":"2026030701434641988_j_auto-2025-0107_ref_062","unstructured":"H. Lin, P. Wang, J. Schneider, and G. Shi, \u201cTD-M(PC)2: Improving temporal difference MPC through policy constraint,\u201d arXiv:2502.03550, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_063","unstructured":"Y. Zhang, S. Yang, T. Ohtsuka, C. Jones, and J. Boedecker, \u201cLatent linear quadratic regulator for robotic control tasks,\u201d arXiv:2407.11107, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_064","unstructured":"M. Watter, J. T. Springenberg, J. Boedecker, and M. Riedmiller, \u201cEmbed to control: A locally linear latent dynamics model for control from raw images,\u201d arXiv:1506.07365, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_065","unstructured":"M. Wang, R. Yang, X. Chen, H. Sun, M. Fang, and G. Montana, \u201cGoplan: Goal-conditioned offline reinforcement learning by planning with learned models,\u201d arXiv:2310.20025, 2023."},{"key":"2026030701434641988_j_auto-2025-0107_ref_066","unstructured":"Y. LeCun, \u201cA path towards autonomous machine intelligence version 0.9. 2, 2022-06-27,\u201d Open Review, vol.\u00a062, no.\u00a01, pp.\u00a01\u201362, 2022."},{"key":"2026030701434641988_j_auto-2025-0107_ref_067","doi-asserted-by":"crossref","unstructured":"M. Assran et al.., \u201cSelf-supervised learning from images with a joint-embedding predictive architecture,\u201d arXiv:2301.08243, 2023.","DOI":"10.1109\/CVPR52729.2023.01499"},{"key":"2026030701434641988_j_auto-2025-0107_ref_068","unstructured":"M. Assran et al.., \u201cV-JEPA 2: Self-supervised video models enable understanding, prediction and planning,\u201d arXiv:2506.09985, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_069","unstructured":"Z. Fei, M. Fan, and J. Huang, \u201cA-jepa: Joint-embedding predictive architecture can listen,\u201d arXiv:2311.15830, 2024."},{"key":"2026030701434641988_j_auto-2025-0107_ref_070","unstructured":"G. Zhou, H. Pan, Y. LeCun, and L. Pinto, \u201cDINO-WM: World models on pre-trained visual features enable zero-shot planning,\u201d arXiv:2411.04983, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_071","unstructured":"M. Oquab et al.., \u201cDINOv2: Learning robust visual features without supervision,\u201d Trans. Mach. Learn. Res., pp.\u00a01\u201331, 2024."},{"key":"2026030701434641988_j_auto-2025-0107_ref_072","unstructured":"A. Vujinovi\u0107 and A. Kova\u010devi\u0107, \u201cACT-JEPA: Novel joint-embedding predictive architecture for efficient policy representation learning,\u201d arXiv:2501.14622, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_073","unstructured":"V. Sobal, W. Zhang, K. Cho, R. Balestriero, T. G. J. Rudner, and Y. LeCun, \u201cLearning from reward-free offline data: A case for planning with latent dynamics models,\u201d arXiv:2502.14819, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_074","doi-asserted-by":"crossref","unstructured":"T. Kenneweg, P. Kenneweg, and B. Hammer, \u201cJEPA for RL: Investigating joint-embedding predictive architectures for reinforcement learning,\u201d arXiv:2504.16591, 2025.","DOI":"10.14428\/esann\/2025.ES2025-19"},{"key":"2026030701434641988_j_auto-2025-0107_ref_075","unstructured":"R. Balestriero and Y. LeCun, \u201cLearning by reconstruction produces uninformative features for perception,\u201d arXiv:2402.11337, 2024."},{"key":"2026030701434641988_j_auto-2025-0107_ref_076","unstructured":"E. Littwin et al.., \u201cHow JEPA avoids noisy features: The implicit bias of deep linear self distillation networks,\u201d arXiv:2407.03475, 2024."},{"key":"2026030701434641988_j_auto-2025-0107_ref_077","doi-asserted-by":"crossref","unstructured":"S. Mo and S. Tong, \u201cConnecting joint-embedding predictive architecture with contrastive self-supervised learning,\u201d arXiv:2410.19560, 2024.","DOI":"10.52202\/079017-0077"},{"key":"2026030701434641988_j_auto-2025-0107_ref_078","unstructured":"J. Yu et al.., \u201cWhy and how auxiliary tasks improve JEPA representations,\u201d arXiv:2509.12249, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_079","doi-asserted-by":"crossref","unstructured":"J. Ulmen, G. Sundaram, and D. G\u00f6rges, \u201cLearning state-space models of dynamic systems from arbitrary data using joint embedding predictive architectures,\u201d arXiv:2508.10489, 2025.","DOI":"10.1016\/j.ifacol.2025.10.190"},{"key":"2026030701434641988_j_auto-2025-0107_ref_080","unstructured":"M. Hartman and L. R. Varshney, \u201cSparseJEPA: Sparse representation learning of joint embedding predictive architectures,\u201d arXiv:2504.16140, 2025."},{"key":"2026030701434641988_j_auto-2025-0107_ref_081","unstructured":"A. Zheng, Evaluating Machine Learning Models: A Beginner\u2019s Guide to Key Concepts and Pitfalls, Sebastopol, CA, O\u2019Reilly Media, 2015."},{"key":"2026030701434641988_j_auto-2025-0107_ref_082","unstructured":"J. Eisenstein, Introduction to Natural Language Processing, Cambridge, MA, MIT Press, 2019."},{"key":"2026030701434641988_j_auto-2025-0107_ref_083","unstructured":"W. Zhu, \u201cClassification of mnist handwritten digit database using neural network,\u201d Proc. Res. School Comput. Sci. Aust. Natl. Univ., Acton, ACT, vol.\u00a02601, 2018."},{"key":"2026030701434641988_j_auto-2025-0107_ref_084","unstructured":"G. Brockman et al.., \u201cOpenAI gym,\u201d arXiv:1606.01540, 2016."}],"container-title":["at - Automatisierungstechnik"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.degruyterbrill.com\/document\/doi\/10.1515\/auto-2025-0107\/xml","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.degruyterbrill.com\/document\/doi\/10.1515\/auto-2025-0107\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,7]],"date-time":"2026-03-07T01:44:13Z","timestamp":1772847853000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.degruyterbrill.com\/document\/doi\/10.1515\/auto-2025-0107\/html"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,1]]},"references-count":84,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,3,9]]},"published-print":{"date-parts":[[2026,3,26]]}},"alternative-id":["10.1515\/auto-2025-0107"],"URL":"https:\/\/doi.org\/10.1515\/auto-2025-0107","relation":{},"ISSN":["0178-2312","2196-677X"],"issn-type":[{"value":"0178-2312","type":"print"},{"value":"2196-677X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,1]]}}}