The DOI matches its Crossref record; 63 of the 63 references it cites were found in a registry or an index.
Liquid Time-constant Networks
Open accessAbstract
We introduce a new class of time-continuous recurrent neural network models. Instead of declaring a learning system's dynamics by implicit nonlinearities, we construct networks of linear first-order dynamical systems modulated via nonlinear interlinked gates. The resulting models represent dynamical systems with varying (i.e., liquid) time-constants coupled to their hidden state, with outputs being computed by numerical differential equation solvers. These neural networks exhibit stable and bounded behavior, yield superior expressivity within the family of neural ordinary differential equations, and give rise to improved performance on time-series prediction tasks. To demonstrate these properties, we first take a theoretical approach to find bounds over their dynamics, and compute their expressive power by the trajectory length measure in a latent trajectory space. We then conduct a series of time-series prediction experiments to manifest the approximation capability of Liquid Time-Constant Networks (LTCs) compared to classical and modern RNNs.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
Learning representations by back-propagating errors
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986) · Nature · doi:10.1038/323533a0 · Cited by 31,892
UCI Machine Learning Repository
Asuncion, A. (2007) · Medical Entomology and Zoology · Cited by 25,410
Multilayer feedforward networks are universal approximators
Hornik, K., Stinchcombe, M., & White, H. (1989) · Neural Networks · doi:10.1016/0893-6080(89)90020-8 · Cited by 21,728
Approximation by superpositions of a sigmoidal function
Cybenko, G. (1989) · Mathematics of Control, Signals, and Systems · doi:10.1007/bf02551274 · Cited by 13,883
Multilayer feedforward networks are universal approximators
HornikK., StinchcombeM., & WhiteH. (1989) · Neural Networks · doi:10.5555/70405.70408 · Cited by 9,081
MuJoCo: A physics engine for model-based control
Todorov, E., Erez, T., & Tassa, Y. (2012) · 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems · doi:10.1109/iros.2012.6386109 · Cited by 4,767
Friston, K. J., Harrison, L., & Penny, W. (2003) · NeuroImage · doi:10.1016/s1053-8119(03)00202-7 · Cited by 4,583
On the difficulty of training recurrent neural networks
Pascanu, R., Mikolov, T., & Bengio, Y. (2012) · arXiv · doi:10.48550/arxiv.1211.5063 · Cited by 4,310
On the approximate realization of continuous mappings by neural networks
Funahashi, K.-I. (1989) · Neural Networks · doi:10.1016/0893-6080(89)90003-8 · Cited by 4,206
Mathematical Theory of Optimal Processes
Pontryagin, L. S. (2018) · Routledge eBooks · doi:10.1201/9780203749319 · Cited by 3,713
A family of embedded Runge-Kutta formulae
Dormand, J. R., & Prince, P. J. (1980) · Journal of Computational and Applied Mathematics · doi:10.1016/0771-050x(80)90013-3 · Cited by 3,634
Numerical Recipes 3rd Edition: The Art of Scientific Computing
Press, W. H., Teukolsky, S. A., Vetterling, W. T., & Flannery, B. P. (2007) · Cambridge University Press eBooks · Cited by 3,328
Differential Equations, Dynamical Systems, and Linear Algebra
(1974) · Pure and Applied Mathematics · doi:10.1016/s0079-8169(08)x6044-1 · Cited by 3,091
Recurrent Neural Networks for Multivariate Time Series with Missing Values
Che, Z., Purushotham, S., Cho, K., Sontag, D., & Liu, Y. (2018) · Scientific Reports · doi:10.1038/s41598-018-24271-9 · Cited by 2,370
Neural ordinary differential equations
Chen, R. T. Q., Rubanova, Y., Bettencourt, J., & Duvenaud, D. (2018) · arXiv · doi:10.48550/arxiv.1806.07366 · Cited by 2,097
(2013) · RECERCAT (Consorci de Serveis Universitaris de Catalunya) · Cited by 1,959
Computer solution of ordinary differential equations : the initial value problem
Shampine, L. F., & Gordon, M. K. (1975) · CERN Document Server (European Organization for Nuclear Research) · Cited by 1,308
Approximation of dynamical systems by continuous time recurrent neural networks
Funahashi, K.-I., & Nakamura, Y. (1993) · Neural Networks · doi:10.1016/s0893-6080(05)80125-x · Cited by 980
On the Number of Linear Regions of Deep Neural Networks
Montúfar, G., Pascanu, R., Cho, K., & Bengio, Y. (2014) · arXiv · doi:10.48550/arxiv.1402.1869 · Cited by 840
Learning Representations by Back-Propagating Errors
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (2002) · Cognitive Modeling · doi:10.7551/mitpress/1888.003.0013 · Cited by 759
Methods in Neuronal Modeling: From Ions to Networks
Koch, C., & Segev, I. (1998) · Cited by 752
Approximation by superpositions of a sigmoidal function
Cybenko, G. (1992) · Mathematics of Control, Signals, and Systems · doi:10.1007/bf02134016 · Cited by 703
Candanedo, L. M., & Feldheim, V. (2016) · Energy and Buildings · doi:10.1016/j.enbuild.2015.11.071 · Cited by 586
Brockman, G., Cheung, V., Pettersson, L., Schneider, J., Schulman, J., Tang, J., & Zaremba, W. (2016) · arXiv · doi:10.48550/arxiv.1606.01540 · Cited by 582
Recherches quantitatives sur l' excitation electrique des nerfs traitee comme une polarization.
Lapique, L. (1907) · Medical Entomology and Zoology · Cited by 562
A 3(2) pair of Runge - Kutta formulas
Bogacki, P., & Shampine, L. F. (1989) · Applied Mathematics Letters · doi:10.1016/0893-9659(89)90079-7 · Cited by 543
(2016) · Conference on Learning Theory · Cited by 443
Computer Solution of Ordinary Differential Equations
Lambert, J. D. (1976) · The Computer Journal · doi:10.1093/comjnl/19.2.155 · Cited by 341
Latent ODEs for Irregularly-Sampled Time SeriesPreprint
Rubanova, Y., Chen, R. T. Q., & Duvenaud, D. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.03907 · Cited by 313
Neural circuit policies enabling auditable autonomy
Lechner, M., Hasani, R., Amini, A., Henzinger, T. A., Rus, D., & Grosu, R. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-00237-3 · Cited by 308
Self-citationRecurrent Neural Networks Are Universal Approximators
Schäfer, A. M., & Zimmermann, H. G. (2006) · Lecture Notes in Computer Science · doi:10.1007/11840817_66 · Cited by 297
Analysis and implementation of TR-BDF2
Hosea, M. E., & Shampine, L. F. (1996) · Applied Numerical Mathematics · doi:10.1016/0168-9274(95)00115-8 · Cited by 291
How to Start Training: The Effect of Initialization and ArchitecturePreprint
Hanin, B., & Rolnick, D. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.01719 · Cited by 285
Exponential expressivity in deep neural networks through transient chaos
Poole, B., Lahiri, S., Raghu, M., Sohl‐Dickstein, J., & Ganguli, S. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.05340 · Cited by 270
Bounding and Counting Linear Regions of Deep Neural NetworksPreprint
Serra, T., Tjandraatmadja, C., & Ramalingam, S. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1711.02114 · Cited by 269
ANODE: Unconditionally Accurate Memory-Efficient Gradients for Neural ODEs
Gholaminejad, A., Keutzer, K., & Biros, G. (2019) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/103 · Cited by 234
Causality for Machine Learning
Schölkopf, B. (2022) · Probabilistic and Causal Inference · doi:10.1145/3501714.3501755 · Cited by 209
The Power of Depth for Feedforward Neural Networks
Eldan, R., & Shamir, O. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1512.03965 · Cited by 200
Learning Long-Term Dependencies in Irregularly-Sampled Time SeriesPreprint
Lechner, M., & Hasani, R. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.04418 · Cited by 176
Self-citationEntropy and mutual information in models of deep neural networks*
Gabrié, M., Manoel, A., Luneau, C., Barbier, J., Macris, N., Krząkała, F., & Zdeborová, L. (2019) · Journal of Statistical Mechanics Theory and Experiment · doi:10.1088/1742-5468/ab3430 · Cited by 169
Complexity of Linear Regions in Deep NetworksPreprint
Hanin, B., & Rolnick, D. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1901.09021 · Cited by 160
Adaptive Checkpoint Adjoint Method for Gradient Estimation in Neural ODE
Zhuang, J., Dvornek, N. C., Li, X., Tatikonda, S., Papademetris, X., & Duncan, J. S. (2020) · PubMed · doi:10.48550/arxiv.2006.02493 · Cited by 159
Wicks, S. R., Roehrig, C. J., & Rankin, C. H. (1996) · The Journal of Neuroscience · doi:10.1523/jneurosci.16-12-04017.1996 · Cited by 156
On the number of response regions of deep feed forward networks with piece-wise linear activations
Pascanu, R., Montufar, G., & Bengio, Y. (2013) · arXiv · doi:10.48550/arxiv.1312.6098 · Cited by 153
Neural Jump Stochastic Differential EquationsPreprint
Jia, J., & Benson, A. R. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.10403 · Cited by 151
On Robustness of Neural Ordinary Differential EquationsPreprint
Yan, H., Du, J., Tan, V. Y. F., & Feng, J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1910.05513 · Cited by 142
(2004) · Elsevier eBooks · doi:10.1016/b978-012264841-0/50054-8 · Cited by 95
ANODE: Unconditionally Accurate Memory-Efficient Gradients for Neural ODEsPreprint
Gholami, A., Keutzer, K., & Biros, G. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.10298 · Cited by 85
Discrete Event, Continuous Time RNNsPreprint
Mozer, M. C., Kazakov, D. V., & Lindsey, R. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1710.04110 · Cited by 79
Forecasting skewed biased stochastic ozone days: analyses, solutions and beyond
Zhang, K., & Fan, W. (2007) · Knowledge and Information Systems · doi:10.1007/s10115-007-0095-1 · Cited by 73
SNODE: Spectral Discretization of Neural ODEs for System IdentificationPreprint
Quaglino, A., Gallieri, M., Masci, J., & Koutník, J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1906.07038 · Cited by 68
Penny, W. D., Ghahramani, Z., & Friston, K. J. (2005) · Philosophical Transactions of the Royal Society B Biological Sciences · doi:10.1098/rstb.2005.1642 · Cited by 67
Designing Worm-inspired Neural Networks for Interpretable Robotic Control
Lechner, M., Hasani, R., Zimmer, M., Henzinger, T. A., & Grosu, R. (2019) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2019.8793840 · Cited by 66
Self-citationTowards Robust, Locally Linear Deep NetworksPreprint
Lee, G.-H., Alvarez-Melis, D., & Jaakkola, T. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.03207 · Cited by 42
The Natural Lottery Ticket Winner: Reinforcement Learning with Ordinary Neural Circuits
Hasani, R., Lechner, M., Amini, A., Rus, D. L., & Grosu, R. (2020) · doi:10.34726/241 · Cited by 34
Self-citationLechner, M., Hasani, R., Rus, D. L., & Grosu, R. (2020) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra40945.2020.9196608 · Cited by 31
Self-citationResponse Characterization for Auditing Cell Dynamics in Long Short-term Memory Networks
Hasani, R., Amini, A., Lechner, M., Naser, F., Grosu, R., & Rus, D. L. (2019) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2019.8851954 · Cited by 21
Self-citationDelPreto, J., Salazar-Gómez, A. F., Gil, S., Hasani, R., Guenther, F. H., & Rus, D. L. (2018) · Robotics: Science and Systems · doi:10.15607/rss.2018.xiv.063 · Cited by 18
Self-citationInterpretable recurrent neural networks in continuous-time control environments
Hasani, R. (2020) · reposiTUm (TU Wien) · doi:10.34726/hss.2020.78942 · Cited by 13
Self-citationNon-Associative Learning Representation in the Nervous System of the Nematode Caenorhabditis elegansPreprint
Hasani, R., Fuchs, M., Beneder, V., & Grosu, R. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1703.06264 · Cited by 12
Self-citationRe-purposing Compact Neuronal Circuit Policies to Govern Reinforcement Learning Tasks.Preprint
Hasani, R., Lechner, M., Amini, A., Rus, D. L., & Grosu, R. (2018) · arXiv (Cornell University) · Cited by 6
Self-citation
Cited by
Neural circuit policies enabling auditable autonomy
Lechner, M., Hasani, R., Amini, A., Henzinger, T. A., Rus, D., & Grosu, R. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-00237-3 · Cited by 308
Self-citationLatent Imagination Facilitates Zero-Shot Transfer in Autonomous Racing
Brunnbauer, A., Berducci, L., Brandstatter, A., Lechner, M., Hasani, R., Rus, D., & Grosu, R. (2022) · 2022 International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra46639.2022.9811650 · Cited by 36
Self-citation
Metadata
- Authors
- Hasani, R., Lechner, M., Amini, A., Rus, D., & Grosu, R.
- Venue
- Proceedings of the AAAI Conference on Artificial Intelligence
- Published
- 2021-05-18
- Volume
- 35
- Issue
- 9
- Pages
- 7657-7666
- ISSN
- 2159-5399 (ISSN-L)
- 2374-3468
- OpenAlex
- W3033739625
- Cited by
- 312
- References
- 63
- FWCI
- 15.30
- Type
- conference paper
- Language
- en
- Open access
- diamond
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.