{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T16:18:00Z","timestamp":1784564280568,"version":"3.55.0"},"reference-count":79,"publisher":"Springer Science and Business Media LLC","issue":"11","license":[{"start":{"date-parts":[[2022,11,15]],"date-time":"2022-11-15T00:00:00Z","timestamp":1668470400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2022,11,15]],"date-time":"2022-11-15T00:00:00Z","timestamp":1668470400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Nat Mach Intell"],"abstract":"<jats:title>Abstract<\/jats:title><jats:p>Continuous-time neural networks are a class of machine learning systems that can tackle representation learning on spatiotemporal decision-making tasks. These models are typically represented by continuous differential equations. However, their expressive power when they are deployed on computers is bottlenecked by numerical differential equation solvers. This limitation has notably slowed down the scaling and understanding of numerous natural physical phenomena such as the dynamics of nervous systems. Ideally, we would circumvent this bottleneck by solving the given dynamical system in closed form. This is known to be intractable in general. Here, we show that it is possible to closely approximate the interaction between neurons and synapses\u2014the building blocks of natural and artificial neural networks\u2014constructed by liquid time-constant networks efficiently in closed form. To this end, we compute a tightly bounded approximation of the solution of an integral appearing in liquid time-constant dynamics that has had no known closed-form solution so far. This closed-form solution impacts the design of continuous-time and continuous-depth neural models. For instance, since time appears explicitly in closed form, the formulation relaxes the need for complex numerical solvers. Consequently, we obtain models that are between one and five orders of magnitude faster in training and inference compared with differential equation-based counterparts. More importantly, in contrast to ordinary differential equation-based continuous networks, closed-form networks can scale remarkably well compared with other deep learning instances. Lastly, as these models are derived from liquid networks, they show good performance in time-series modelling compared with advanced recurrent neural network models.<\/jats:p>","DOI":"10.1038\/s42256-022-00556-7","type":"journal-article","created":{"date-parts":[[2022,11,15]],"date-time":"2022-11-15T17:03:19Z","timestamp":1668531799000},"page":"992-1003","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":182,"title":["Closed-form continuous-time neural networks"],"prefix":"10.1038","volume":"4","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9889-5222","authenticated-orcid":false,"given":"Ramin","family":"Hasani","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mathias","family":"Lechner","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexander","family":"Amini","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3229-6665","authenticated-orcid":false,"given":"Lucas","family":"Liebenwein","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aaron","family":"Ray","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Max","family":"Tschaikowski","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1036-9173","authenticated-orcid":false,"given":"Gerald","family":"Teschl","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Daniela","family":"Rus","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2022,11,15]]},"reference":[{"key":"556_CR1","doi-asserted-by":"crossref","unstructured":"Hasani, R., Lechner, M., Amini, A., Rus, D. & Grosu, R. Liquid time-constant networks. In Proc. of AAAI Conference on Artificial Intelligence 35(9), 7657\u20137666 (AAAI, 2021).","DOI":"10.1609\/aaai.v35i9.16936"},{"key":"556_CR2","unstructured":"Chen, T. Q., Rubanova, Y., Bettencourt, J. & Duvenaud, D. K. Neural ordinary differential equations. In Proc. of Advances in Neural Information Processing Systems (Eds. Bengio, S. et al.) 6571\u20136583 (NeurIPS, 2018)."},{"key":"556_CR3","unstructured":"Grathwohl, W., Chen, R. T., Bettencourt, J., Sutskever, I. & Duvenaud, D. Ffjord: free-form continuous dynamics for scalable reversible generative models. In International Conference on Learning Representations (2018). https:\/\/openreview.net\/forum?id=rJxgknCcK7"},{"key":"556_CR4","unstructured":"Dupont, E., Doucet, A. & Teh, Y. W. Augmented neural ODEs. In Proc. of Advances in Neural Information Processing Systems (Eds. Wallach, H. et al.) 3134\u20133144 (NeurIPS, 2019)."},{"key":"556_CR5","doi-asserted-by":"crossref","unstructured":"Yang, G. et al. Pointflow: 3D point cloud generation with continuous normalizing flows. In Proc. of the IEEE\/CVF International Conference on Computer Vision 4541\u20134550 (IEEE, 2019).","DOI":"10.1109\/ICCV.2019.00464"},{"key":"556_CR6","unstructured":"Liebenwein, L., Hasani, R., Amini, A. & Daniela, R. Sparse flows: pruning continuous-depth models. In Proc. of Advances in Neural Information Processing Systems (Eds. Ranzato, M. et al.) 22628\u201322642 (NeurIPS, 2021)."},{"key":"556_CR7","unstructured":"Rubanova, Y., Chen, R. T. & Duvenaud, D. Latent Neural ODEs for irregularly-sampled time series. In Proc. of Advances in Neural Information Processing Systems (Eds. Wallach, H. et al.) 32 (NeurIPS, 2019)."},{"key":"556_CR8","doi-asserted-by":"crossref","unstructured":"Gholami, A., Keutzer, K. & Biros, G. ANODE: unconditionally accurate memory-efficient gradients for neural ODEs. In Proceedings of the 28th International Joint Conference on Artificial Intelligence 730\u2013736 (IJCAI, 2019).","DOI":"10.24963\/ijcai.2019\/103"},{"key":"556_CR9","unstructured":"Lechner, M. & Hasani, R. Learning long-term dependencies in irregularly-sampled time series. Preprint at https:\/\/arxiv.org\/abs\/2006.04418 (2020)."},{"key":"556_CR10","doi-asserted-by":"publisher","first-page":"67","DOI":"10.1016\/0771-050X(81)90010-3","volume":"7","author":"PJ Prince","year":"1981","unstructured":"Prince, P. J. & Dormand, J. R. High order embedded Runge\u2013Kutta formulae. J. Comput. Appl. Math. 7, 67\u201375 (1981).","journal-title":"J. Comput. Appl. Math."},{"key":"556_CR11","doi-asserted-by":"publisher","first-page":"686","DOI":"10.1016\/j.jcp.2018.10.045","volume":"378","author":"M Raissi","year":"2019","unstructured":"Raissi, M., Perdikaris, P. & Karniadakis, G. E. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J. Comput. Phys. 378, 686\u2013707 (2019).","journal-title":"J. Comput. Phys."},{"key":"556_CR12","unstructured":"Massaroli, S., Poli, M., Park, J., Yamashita, A. & Asma, H. Dissecting neural ODEs. In Proc. of 33th Conference on Neural Information Processing Systems (Eds. Larochelle, H. et al.) (NeurIPS, 2020)."},{"key":"556_CR13","first-page":"690","volume":"32","author":"S Bai","year":"2019","unstructured":"Bai, S., Kolter, J. Z. & Koltun, V. Deep equilibrium models. Adv. Neural Inform. Process. Syst. 32, 690\u2013701 (2019).","journal-title":"Adv. Neural Inform. Process. Syst."},{"key":"556_CR14","unstructured":"Finlay, C., Jacobsen, J.-H., Nurbekyan, L. & Oberman, A. M. How to train your neural ODE: the world of Jacobian and kinetic regularization. In International Conference on Machine Learning (Eds. Daum\u00e9 III, H. & Singh, A.) 3154\u20133164 (PMLR, 2020)."},{"key":"556_CR15","unstructured":"Massaroli, S. et al. Stable Neural Flows. Preprint at https:\/\/arxiv.org\/abs\/2003.08063 (2020)."},{"key":"556_CR16","unstructured":"Kidger, P., Chen, R. T. & Lyons, T. \u201cHey, that\u2019s not an ODE\u201d: Faster ODE Adjoints via Seminorms. In Proceedings of the 38th International Conference on Machine Learning (Eds. Meila, M. & Zhang, T.) 139 (PMLR, 2021)."},{"key":"556_CR17","unstructured":"Poli, M. et al. Hypersolvers: toward fast continuous-depth models. In Proc. of Advances in Neural Information Processing Systems (Eds. Larochelle, H.) 21105\u201321117 (NeurIPS, 2020)."},{"key":"556_CR18","doi-asserted-by":"publisher","first-page":"056215","DOI":"10.1103\/PhysRevE.85.056215","volume":"85","author":"J Schumacher","year":"2012","unstructured":"Schumacher, J., Haslinger, R. & Pipa, G. Statistical modeling approach for detecting generalized synchronization. Phys. Rev. E 85, 056215 (2012).","journal-title":"Phys. Rev. E"},{"key":"556_CR19","doi-asserted-by":"publisher","first-page":"57","DOI":"10.3389\/fncom.2013.00057","volume":"7","author":"R Moran","year":"2013","unstructured":"Moran, R., Pinotsis, D. A. & Friston, K. Neural masses and fields in dynamic causal modeling. Front. Comput. Neurosci. 7, 57 (2013).","journal-title":"Front. Comput. Neurosci."},{"key":"556_CR20","doi-asserted-by":"publisher","first-page":"1273","DOI":"10.1016\/S1053-8119(03)00202-7","volume":"19","author":"KJ Friston","year":"2003","unstructured":"Friston, K. J., Harrison, L. & Penny, W. Dynamic causal modelling. Neuroimage 19, 1273\u20131302 (2003).","journal-title":"Neuroimage"},{"key":"556_CR21","doi-asserted-by":"publisher","DOI":"10.1007\/978-1-4684-0392-3","volume-title":"Differential Equations and Dynamical Systems","author":"L Perko","year":"1991","unstructured":"Perko, L. Differential Equations and Dynamical Systems (Springer-Verlag, 1991)."},{"key":"556_CR22","doi-asserted-by":"publisher","first-page":"642","DOI":"10.1038\/s42256-020-00237-3","volume":"2","author":"M Lechner","year":"2020","unstructured":"Lechner, M. et al. Neural circuit policies enabling auditable autonomy. Nat. Mach. Intell. 2, 642\u2013652 (2020).","journal-title":"Nat. Mach. Intell."},{"key":"556_CR23","unstructured":"Hochreiter, S. Untersuchungen zu dynamischen neuronalen netzen. Diploma, Technische Universit\u00e4t M\u00fcnchen 91 (1991)."},{"key":"556_CR24","unstructured":"Vorbach, C., Hasani, R., Amini, A., Lechner, M. & Rus, D. Causal navigation by continuous-time neural networks. In Proc. of Advances in Neural Information Processing Systems (Eds. Ranzato, M. et al.) 12425\u201312440 (NeurIPS, 2021)."},{"key":"556_CR25","doi-asserted-by":"crossref","unstructured":"Hasani, R. et al. Response characterization for auditing cell dynamics in long short-term memory networks. In Proc. of 2019 International Joint Conference on Neural Networks 1\u20138 (IEEE, 2019).","DOI":"10.1109\/IJCNN.2019.8851954"},{"key":"556_CR26","unstructured":"Anguita, D., Ghio, A., Oneto, L., Parra Perez, X. & Reyes Ortiz, J. L. A public domain dataset for human activity recognition using smartphones. In Proc. of the 21st International European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning 437\u2013442 (i6doc, 2013)."},{"key":"556_CR27","doi-asserted-by":"crossref","unstructured":"Todorov, E., Erez, T. & Tassa, Y. MuJoCo: a physics engine for model-based control. In Proc. of 2012 IEEE\/RSJ International Conference on Intelligent Robots and Systems 5026\u20135033 (IEEE, 2012).","DOI":"10.1109\/IROS.2012.6386109"},{"key":"556_CR28","unstructured":"Maas, A. et al. Learning word vectors for sentiment analysis. In Proc. of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies 142\u2013150 (ACM, 2011)."},{"key":"556_CR29","doi-asserted-by":"publisher","first-page":"218","DOI":"10.1038\/s42256-021-00302-5","volume":"3","author":"L Lu","year":"2021","unstructured":"Lu, L., Jin, P., Pang, G., Zhang, Z. & Karniadakis, G. E. Learning nonlinear operators via deeponet based on the universal approximation theorem of operators. Nat. Mach. Intell. 3, 218\u2013229 (2021).","journal-title":"Nat. Mach. Intell."},{"key":"556_CR30","doi-asserted-by":"publisher","first-page":"422","DOI":"10.1038\/s42254-021-00314-5","volume":"3","author":"GE Karniadakis","year":"2021","unstructured":"Karniadakis, G. E. et al. Physics-informed machine learning. Nat. Rev. Phys. 3, 422\u2013440 (2021).","journal-title":"Nat. Rev. Phys."},{"key":"556_CR31","doi-asserted-by":"publisher","first-page":"eabi8605","DOI":"10.1126\/sciadv.abi8605","volume":"7","author":"S Wang","year":"2021","unstructured":"Wang, S., Wang, H. & Perdikaris, P. Learning the solution operator of parametric partial differential equations with physics-informed deeponets. Sci. Adv. 7, eabi8605 (2021).","journal-title":"Sci. Adv."},{"key":"556_CR32","unstructured":"Rezende, D. & Mohamed, S. Variational inference with normalizing flows. In Proc. of International Conference on Machine Learning (Eds. Bach, F. & Blei, D.) 1530\u20131538 (PMLR, 2015)."},{"key":"556_CR33","unstructured":"Gu, A., Goel, K. & Re, C. Efficiently modeling long sequences with structured state spaces. In Proc. of International Conference on Learning Representations (2022). https:\/\/openreview.net\/forum?id=uYLFoz1vlAC"},{"key":"556_CR34","unstructured":"Hasani, R. et al. Liquid structural state-space models. Preprint at https:\/\/arxiv.org\/abs\/2209.12951 (2022)."},{"key":"556_CR35","first-page":"11525","volume":"35","author":"S Grunbacher","year":"2021","unstructured":"Grunbacher, S. et al. On the verification of neural ODEs with stochastic guarantees. Proc. AAAI Conf. Artif. Intell. 35, 11525\u201311535 (2021).","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"556_CR36","unstructured":"Vaswani, A. et al. Attention is all you need. In Proc. of Advances in Neural Information Processing Systems (Eds. Guyon, I. et al.) 5998\u20136008 (NeurIPS, 2017)."},{"key":"556_CR37","doi-asserted-by":"crossref","unstructured":"Lechner, M., Hasani, R., Grosu, R., Rus, D. & Henzinger, T. A. Adversarial training is not ready for robot learning. In 2021 IEEE International Conference on Robotics and Automation (ICRA) 4140\u20134147 (IEEE, 2021).","DOI":"10.1109\/ICRA48506.2021.9561036"},{"key":"556_CR38","doi-asserted-by":"crossref","unstructured":"Brunnbauer, A. et al. Latent imagination facilitates zero-shot transfer in autonomous racing. In 2022 International Conference on Robotics and Automation (ICRA) 7513\u20137520 (IEEE, 2021).","DOI":"10.1109\/ICRA46639.2022.9811650"},{"key":"556_CR39","doi-asserted-by":"crossref","unstructured":"Hasani, R. M., Haerle, D. & Grosu, R. Efficient modeling of complex analog integrated circuits using neural networks. In Proc. of 12th Conference on Ph.D. Research in Microelectronics and Electronics 1\u20134 (IEEE, 2016).","DOI":"10.1109\/PRIME.2016.7519486"},{"key":"556_CR40","doi-asserted-by":"publisher","first-page":"105683","DOI":"10.1016\/j.asoc.2019.105683","volume":"85","author":"G Wang","year":"2019","unstructured":"Wang, G., Ledwoch, A., Hasani, R. M., Grosu, R. & Brintrup, A. A generative neural network model for the quality prediction of work in progress products. Appl. Soft Comput. 85, 105683 (2019).","journal-title":"Appl. Soft Comput."},{"key":"556_CR41","doi-asserted-by":"publisher","first-page":"1303","DOI":"10.1007\/s10514-020-09916-x","volume":"44","author":"J DelPreto","year":"2020","unstructured":"DelPreto, J. et al. Plug-and-play supervisory control using muscle and brain signals for real-time gesture and error detection. Auton. Robots 44, 1303\u20131322 (2020).","journal-title":"Auton. Robots"},{"key":"556_CR42","unstructured":"Hasani, R. Interpretable Recurrent Neural Networks in Continuous-Time Control Environments. PhD dissertation, Technische Univ. Wien (2020)."},{"key":"556_CR43","unstructured":"Rudin, W. Principles of Mathematical Analysis, 3rd edn. (McGraw-Hill, 1976)."},{"key":"556_CR44","doi-asserted-by":"publisher","first-page":"1735","DOI":"10.1162\/neco.1997.9.8.1735","volume":"9","author":"S Hochreiter","year":"1997","unstructured":"Hochreiter, S. & Schmidhuber, J. Long short-term memory. Neural Comput. 9, 1735\u20131780 (1997).","journal-title":"Neural Comput."},{"key":"556_CR45","unstructured":"Chung, J., Gulcehre, C., Cho, K. & Bengio, Y. Empirical evaluation of gated recurrent neural networks on sequence modeling. Preprint at https:\/\/arxiv.org\/abs\/1412.3555 (2014)."},{"key":"556_CR46","unstructured":"Shukla, S. N. & Marlin, B. Interpolation\u2013prediction networks for irregularly sampled time series. In Proc. of International Conference on Learning Representations (2018). https:\/\/openreview.net\/forum?id=r1efr3C9Ym"},{"key":"556_CR47","unstructured":"Horn, M., Moor, M., Bock, C., Rieck, B. & Borgwardt, K. Set functions for time series. In Proc. of International Conference on Machine Learning (Eds. Daum\u00e9 III, H. & Singh, A.) 4353\u20134363 (PMLR, 2020)."},{"key":"556_CR48","doi-asserted-by":"publisher","first-page":"801","DOI":"10.1016\/S0893-6080(05)80125-X","volume":"6","author":"K-i Funahashi","year":"1993","unstructured":"Funahashi, K.-i & Nakamura, Y. Approximation of dynamical systems by continuous time recurrent neural networks. Neural Netw. 6, 801\u2013806 (1993).","journal-title":"Neural Netw."},{"key":"556_CR49","unstructured":"Mozer, M. C., Kazakov, D. & Lindsey, R. V. Discrete event, continuous time RNNs. Preprint at https:\/\/arxiv.org\/abs\/1710.04110 (2017)."},{"key":"556_CR50","unstructured":"Mei, H. & Eisner, J. The neural Hawkes process: a neurally self-modulating multivariate point process. In Proc. of 31st International Conference on Neural Information Processing Systems (Eds. Guyon, I. et al.) 6757\u20136767 (Curran Associates Inc., 2017)."},{"key":"556_CR51","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41598-018-24271-9","volume":"8","author":"Z Che","year":"2018","unstructured":"Che, Z., Purushotham, S., Cho, K., Sontag, D. & Liu, Y. Recurrent neural networks for multivariate time series with missing values. Sci. Rep. 8, 1\u201312 (2018).","journal-title":"Sci. Rep."},{"key":"556_CR52","unstructured":"Neil, D., Pfeiffer, M. & Liu, S.-C. Phased LSTM: accelerating recurrent network training for long or event-based sequences. In Proc. of 30th International Conference on Neural Information Processing Systems (Eds. Lee, D. D. et al.) 3889\u20133897 (Curran Associates Inc., 2016)."},{"key":"556_CR53","doi-asserted-by":"publisher","first-page":"2673","DOI":"10.1109\/78.650093","volume":"45","author":"M Schuster","year":"1997","unstructured":"Schuster, M. & Paliwal, K. K. Bidirectional recurrent neural networks. IEEE Trans. Signal Process. 45, 2673\u20132681 (1997).","journal-title":"IEEE Trans. Signal Process."},{"key":"556_CR54","unstructured":"Voelker, A. R., Kaji\u0107, I. & Eliasmith, C. Legendre memory units: continuous-time representation in recurrent neural networks. In Proceedings of the 33rd International Conference on Neural Information Processing Systems (Eds. Wallach, H. et al.) 15570\u201315579 (ACM, 2019)."},{"key":"556_CR55","unstructured":"Gu, A., Dao, T., Ermon, S., Rudra, A. & R\u00e9, C. Hippo: recurrent memory with optimal polynomial projections. In Proc. of Advances in Neural Information Processing Systems (Eds. Larochelle, H. et al.) 1474\u20131487 (NeurIPS, 2020)."},{"key":"556_CR56","unstructured":"Lezcano-Casado, M. & Mart\u0131nez-Rubio, D. Cheap orthogonal constraints in neural networks: a simple parametrization of the orthogonal and unitary group. In Proc. of International Conference on Machine Learning (Eds. Chaudhuri, K. & Salakhutdinov, R.) 3794\u20133803 (PMLR, 2019)."},{"key":"556_CR57","unstructured":"Rusch, T. K. & Mishra, S. Coupled oscillatory recurrent neural network (coRNN): an accurate and (gradient) stable architecture for learning long time dependencies. In Proc. of International Conference on Learning Representations (2021). https:\/\/openreview.net\/forum?id=F3s69XzWOia"},{"key":"556_CR58","unstructured":"Erichson, N. B., Azencot, O., Queiruga, A., Hodgkinson, L. & Mahoney, M. W. Lipschitz recurrent neural networks. In Proc. of International Conference on Learning Representations (2021). https:\/\/openreview.net\/forum?id=-N7PBXqOUJZ"},{"key":"556_CR59","unstructured":"Brockman, G. et al. OpenAI gym. Preprint at https:\/\/arxiv.org\/abs\/1606.01540 (2016)."},{"key":"556_CR60","doi-asserted-by":"crossref","unstructured":"Lechner, M., Hasani, R., Zimmer, M., Henzinger, T. A. & Grosu, R. Designing worm-inspired neural networks for interpretable robotic control. In Proc. of International Conference on Robotics and Automation 87\u201394 (IEEE, 2019).","DOI":"10.1109\/ICRA.2019.8793840"},{"key":"556_CR61","doi-asserted-by":"publisher","first-page":"3265","DOI":"10.1109\/LRA.2022.3146555","volume":"7","author":"P Tylkin","year":"2022","unstructured":"Tylkin, P. et al. Interpretable autonomous flight via compact visualizable neural circuit policies. IEEE Robot. Autom. Lett. 7, 3265\u20133272 (2022).","journal-title":"IEEE Robot. Autom. Lett."},{"key":"556_CR62","doi-asserted-by":"crossref","unstructured":"Amini, A. et al. Vista 2.0: An open, data-driven simulator for multimodal sensing and policy learning for autonomous vehicles. In 2022 International Conference on Robotics and Automation (ICRA) 2419\u20132426 (IEEE, 2022).","DOI":"10.1109\/ICRA46639.2022.9812276"},{"key":"556_CR63","doi-asserted-by":"crossref","unstructured":"Amini, A. et al. Learning robust control policies for end-to-end autonomous driving from data-driven simulation. IEEE Robot. Autom. Lett. 5, 1143\u20131150 (2020).","DOI":"10.1109\/LRA.2020.2966414"},{"key":"556_CR64","unstructured":"Levine, S. & Koltun, V. Guided policy search. In Proc. of International Conference on Machine Learning (Eds. Dasgupta, S. & McAllester, D.) 1\u20139 (PMLR, 2013)."},{"key":"556_CR65","doi-asserted-by":"crossref","unstructured":"Bojarski, M. et al. VisualBackProp: efficient visualization of CNNs for autonomous driving. In Proc. of IEEE International Conference on Robotics and Automation 1\u20138 (IEEE, 2018).","DOI":"10.1109\/ICRA.2018.8461053"},{"key":"556_CR66","doi-asserted-by":"publisher","first-page":"1229","DOI":"10.1109\/TNNLS.2014.2317880","volume":"25","author":"H Zhang","year":"2014","unstructured":"Zhang, H., Wang, Z. & Liu, D. A comprehensive review of stability analysis of continuous-time recurrent neural networks. IEEE Trans. Neural Netw. Learn. Syst 25, 1229\u20131262 (2014).","journal-title":"IEEE Trans. Neural Netw. Learn. Syst"},{"key":"556_CR67","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s40304-017-0103-z","volume":"5","author":"E Weinan","year":"2017","unstructured":"Weinan, E. A proposal on machine learning via dynamical systems. Commun. Math. Stat. 5, 1\u201311 (2017).","journal-title":"Commun. Math. Stat."},{"key":"556_CR68","unstructured":"Lu, Z., Pu, H., Wang, F., Hu, Z. & Wang, L. The expressive power of neural networks: a view from the width. In Proc. of Advances in Neural Information Processing Systems (Eds. Guyon, I. et al.) 30 (Curran Associates, Inc 2017)."},{"key":"556_CR69","unstructured":"Li, Q., Chen, L., Tai, C. et al. Maximum principle based algorithms for deep learning. J. Mach. Learn. Res. 18, 5998\u20136026 (2018)."},{"key":"556_CR70","doi-asserted-by":"publisher","first-page":"815","DOI":"10.1109\/TSMC.1983.6313075","volume":"5","author":"MA Cohen","year":"1983","unstructured":"Cohen, M. A. & Grossberg, S. Absolute stability of global pattern formation and parallel memory storage by competitive neural networks. IEEE Trans. Syst. Man Cybern. 5, 815\u2013826 (1983).","journal-title":"IEEE Trans. Syst. Man Cybern."},{"key":"556_CR71","unstructured":"Mathieu, E. & Nickel, M. Riemannian continuous normalizing flows. In Proc. of Advances in Neural Information Processing Systems Vol. 33 (eds Larochelle et al.) 2503\u20132515 (Curran Associates, Inc., 2020)."},{"key":"556_CR72","unstructured":"Hodgkinson, L., van der Heide, C., Roosta, F. & Mahoney, M. W. Stochastic normalizing flows. In Proc. of Advances in Neural Information Processing Systems (Eds. Larochelle, H. et al.) 5933\u20135944 (NeurIPS, 2020)."},{"key":"556_CR73","unstructured":"Haber, E., Lensink, K., Treister, E. & Ruthotto, L. IMEXnet a forward stable deep neural network. In Proc. of International Conference on Machine Learning (Eds. Chaudhuri, K. & Salakhutdinov, R.) 2525\u20132534 (PMLR, 2019)."},{"key":"556_CR74","unstructured":"Chang, B., Chen, M., Haber, E. & Chi, E. H. AntisymmetricRNN: a dynamical system view on recurrent neural networks. In International Conference on Learning Representations (2018). https:\/\/openreview.net\/forum?id=ryxepo0cFX"},{"key":"556_CR75","doi-asserted-by":"crossref","unstructured":"Lechner, M., Hasani, R., Rus, D. & Grosu, R. Gershgorin loss stabilizes the recurrent neural network compartment of an end-to-end robot learning scheme. In Proc. of IEEE International Conference on Robotics and Automation 5446\u20135452 (IEEE, 2020).","DOI":"10.1109\/ICRA40945.2020.9196608"},{"key":"556_CR76","doi-asserted-by":"publisher","first-page":"20170379","DOI":"10.1098\/rstb.2017.0379","volume":"373","author":"P Gleeson","year":"2018","unstructured":"Gleeson, P., Lung, D., Grosu, R., Hasani, R. & Larson, S. D. c302: a multiscale framework for modelling the nervous system of Caenorhabditis elegans. Philos.Trans. R. Soc. B 373, 20170379 (2018).","journal-title":"Philos.Trans. R. Soc. B"},{"key":"556_CR77","unstructured":"Li, X., Wong, T.-K. L., Chen, R. T. & Duvenaud, D. Scalable gradients for stochastic differential equations. In Proc. of International Conference on Artificial Intelligence and Statistics 3870\u20133882 (PMLR, 2020)."},{"key":"556_CR78","unstructured":"Shukla, S. N. & Marlin, B. M. Multi-time attention networks for irregularly sampled time series. In International Conference on Learning Representations (2020). https:\/\/openreview.net\/forum?id=4c0J6lwQ4_"},{"key":"556_CR79","doi-asserted-by":"crossref","unstructured":"Xiong, Y. et al. Nystr\u00f6mformer: a Nystr\u00f6m-based algorithm for approximating self-attention. In Proceedings of the AAAI Conference on Artificial Intelligence Vol. 35, No. 16, pp. 14138\u201314148 (AAAI, 2021).","DOI":"10.1609\/aaai.v35i16.17664"}],"updated-by":[{"DOI":"10.1038\/s42256-022-00597-y","type":"correction","label":"Correction","source":"publisher","updated":{"date-parts":[[2022,11,29]],"date-time":"2022-11-29T00:00:00Z","timestamp":1669680000000}}],"container-title":["Nature Machine Intelligence"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s42256-022-00556-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s42256-022-00556-7","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s42256-022-00556-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,10,8]],"date-time":"2024-10-08T07:25:37Z","timestamp":1728372337000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s42256-022-00556-7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,11,15]]},"references-count":79,"journal-issue":{"issue":"11","published-online":{"date-parts":[[2022,11]]}},"alternative-id":["556"],"URL":"https:\/\/doi.org\/10.1038\/s42256-022-00556-7","relation":{"correction":[{"id-type":"doi","id":"10.1038\/s42256-022-00597-y","asserted-by":"object"}]},"ISSN":["2522-5839"],"issn-type":[{"value":"2522-5839","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,11,15]]},"assertion":[{"value":"23 March 2022","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"5 October 2022","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"15 November 2022","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"29 November 2022","order":4,"name":"change_date","label":"Change Date","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"Correction","order":5,"name":"change_type","label":"Change Type","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"A Correction to this paper has been published:","order":6,"name":"change_details","label":"Change Details","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"https:\/\/doi.org\/10.1038\/s42256-022-00597-y","URL":"https:\/\/doi.org\/10.1038\/s42256-022-00597-y","order":7,"name":"change_details","label":"Change Details","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interest.","order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}]}}