CoChat

Flexible multitask computation in recurrent networks utilizes shared dynamical motifs

Open access
Driscoll, Laura N., Shenoy, Krishna, Sussillo, David · 2024 · Nature Neuroscience
Cited by 214

The DOI matches its Crossref record; 18 of the 18 references it cites were found in a registry or an index.

Abstract

Flexible computation is a hallmark of intelligent behavior. However, little is known about how neural networks contextually reconfigure for different computations. In the present work, we identified an algorithmic neural substrate for modular computation through the study of multitasking artificial recurrent neural networks. Dynamical systems analyses revealed learned computational strategies mirroring the modular subtask structure of the training task set. Dynamical motifs, which are recurring patterns of neural activity that implement specific computations through dynamics, such as attractors, decision boundaries and rotations, were reused across tasks. For example, tasks requiring memory of a continuous circular variable repurposed the same ring attractor. We showed that dynamical motifs were implemented by clusters of units when the unit activation function was restricted to be positive. Cluster lesions caused modular performance deficits. Motifs were reconfigured for fast transfer learning after an initial phase of learning. This work establishes dynamical motifs as a fundamental unit of compositional computation, intermediate between neuron and network. As whole-brain studies simultaneously record activity from multiple specialized systems, the dynamical motif framework will guide questions about specialization and generalization.

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.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41593-024-01668-6

Cites

18 of 18 references verified0 citation errors · 0 not checkable

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.

A few free checks a day · papers already verified open instantly · results are public

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

Show all 18 references

Research works printed but not linked: 44

  1. Cole, M. W. et al. Multi-task connectivity reveals flexible hubs for adaptive task control. Nat. Neurosci. 16, 1348–1355 (2013). · DOI · Paper page
  2. Yang, G. R., Cole, M. W. & Rajan, K. How to study the neural mechanisms of multiple tasks. Curr. Opin. Behav. Sci. 29, 134–143 (2019). · DOI
  3. Frankland, S. M. & Greene, J. D. Concepts and compositionality: in search of the brain’s language of thought. Annu. Rev. Psychol. 71, 273–303 (2020). · DOI · Paper page
  4. Reverberi, C., Görgen, K. & Haynes, J.-D. Compositionality of rule representations in human prefrontal cortex. Cereb. Cortex 22, 1237–1246 (2012). · DOI · Paper page
  5. Barak, O., Sussillo, D., Romo, R., Tsodyks, M. & Abbott, L. F. From fixed points to chaos: three models of delayed discrimination. Prog. Neurobiol. 103, 214–222 (2013). · DOI · Paper page
  6. Mante, V., Sussillo, D., Shenoy, K. V. & Newsome, W. T. Context-dependent computation by recurrent dynamics in prefrontal cortex. Nature 503, 78–84 (2013). · DOI · Paper page
  7. Carnevale, F., de Lafuente, V., Romo, R., Barak, O. & Parga, N. Dynamic control of response criterion in premotor cortex during perceptual detection under temporal uncertainty. Neuron 86, 1067–1077 (2015). · DOI
  8. Sussillo, D., Churchland, M. M., Kaufman, M. T. & Shenoy, K. V. A neural network that finds a naturalistic solution for the production of muscle activity. Nat. Neurosci. 18, 1025–1033 (2015). · DOI · Paper page
  9. Remington, E. D., Narain, D., Hosseini, E. A. & Jazayeri, M. Flexible sensorimotor computations through rapid reconfiguration of cortical dynamics. Neuron 98, 1005–1019 (2018). · DOI · Paper page
  10. Mastrogiuseppe, F. & Ostojic, S. Linking connectivity, dynamics, and computations in low-rank recurrent neural networks. Neuron 99, 609–623 (2018). · DOI · Paper page
  11. Russo, A. A. et al. Motor cortex embeds muscle-like commands in an untangled population response. Neuron 97, 953–966 (2018). · DOI · Paper page
  12. Pinto, L. et al. Task-dependent changes in the large-scale dynamics and necessity of cortical regions. Neuron 104, 810–824 (2019). · DOI · Paper page
  13. Duncker, L., Driscoll, L., Shenoy, K. V., Sahani, M. & Sussillo, D. Organizing recurrent network dynamics by task-computation to enable continual learning. In Proc. 34th Conference on Neural Information Processing Systems https://proceedings.neurips.cc/paper/2020/file/a576eafbce762079f7d1f77fca1c5cc2-Paper.pdf (NeurIPS, 2020).
  14. Masse, N. Y., Grant, G. D. & Freedman, D. J. Alleviating catastrophic forgetting using context-dependent gating and synaptic stabilization. Proc. Natl Acad. Sci. USA 115, E10467–E10475 (2018). · DOI · Paper page
  15. Riveland, R. & Pouget, A. Natural language instructions induce compositional generalization in networks of neurons. Nat. Neurosci. 27, 988–999 (2024). · DOI
  16. Yang, G. R., Ganichev, I., Wang, X.-J., Shlens, J. & Sussillo, D. A dataset and architecture for visual reasoning with a working memory. In 2018 Conference on Cognitive Computational Neuroscience https://doi.org/10.32470/ccn.2018.1075-0 (Cognitive Computational Neuroscience, 2018). · doi:10.32470/ccn.2018.1075-0
  17. Márton, C. D., Gagnon, L., Lajoie, G. & Rajan, K. Efficient and robust multi-task learning in the brain with modular latent primitives. Preprint at https://arxiv.org/abs/2105.14108 (2021). · arXiv
  18. Strogatz, S. H. & Strogatz, R. Nonlinear Dynamics and Chaos: With Applications to Physics, Biology, Chemistry, and Engineering 2nd edn (Westview Press, 1994). · DOI · Paper page
  19. Golub, M. & Sussillo, D. FixedPointFinder: a Tensorflow toolbox for identifying and characterizing fixed points in recurrent neural networks. J. Open Source Softw. 3, 1003 (2018). · DOI
  20. Kaufman, M. T., Churchland, M. M., Ryu, S. I. & Shenoy, K. V. Cortical activity in the null space: permitting preparation without movement. Nat. Neurosci. 17, 440–448 (2014). · DOI · Paper page
  21. Pagan, M. et al. A new theoretical framework jointly explains behavioral and neural variability across subjects performing flexible decision-making. Preprint at bioRxiv https://doi.org/10.1101/2022.11.28.518207 (2022). · DOI
  22. Baroni, M. Linguistic generalization and compositionality in modern artificial neural networks. Phil. Trans. R. Soc. Lond. B 375, 20190307 (2020). · DOI · Paper page
  23. Morehead, J. R., Qasim, S. E., Crossley, M. J. & Ivry, R. Savings upon re-aiming in visuomotor adaptation. J. Neurosci. 35, 14386–14396 (2015). · DOI · Paper page
  24. Golub, M. D. et al. Learning by neural reassociation. Nat. Neurosci. 21, 607–616 (2018). · DOI · Paper page
  25. Han, X. et al. Pre-trained models: past, present and future. AI Open 2, 225–250 (2021). · DOI · Paper page
  26. Maheswaranathan, N., Williams, A. H., Golub, M. D., Ganguli, S. & Sussillo, D. Universality and individuality in neural dynamics across large populations of recurrent networks. In Proc. 33rd Conference on Neural Information Processing Systems https://proceedings.neurips.cc/paper_files/paper/2019/file/07a4e20a7bbeeb7a736682b26b16ebe8-Paper.pdf (NeurIPS, 2019).
  27. Turner, E. & Barak, O. The simplicity bias in multi-task RNNs: shared attractors, reuse of dynamics, and geometric representation. In Proc. 37th Conference on Neural Information Processing Systems https://proceedings.neurips.cc/paper_files/paper/2023/file/50d6dbc809b0dc96f7f1090810537acc-Paper-Conference.pdf (NeurIPS, 2023).
  28. Goudar, V., Peysakhovich, B., Freedman, D. J., Buffalo, E. A. & Wang, X.-J. Schema formation in a neural population subspace underlies learning-to-learn in flexible sensorimotor problem-solving. Nat. Neurosci. 26, 879–890 (2023). · DOI
  29. Chen, G., Kang, B., Lindsey, J., Druckmann, S. & Li, N. Modularity and robustness of frontal cortical networks. Cell 184, 3717–3730 (2021). · DOI
  30. Wang, J., Narain, D., Hosseini, E. A. & Jazayeri, M. Flexible timing by temporal scaling of cortical responses. Nat. Neurosci. 21, 102–110 (2018). · DOI · Paper page
  31. Jaeger, H. in Dynamics, Synergetics, Autonomous Agents Vol. 8 (eds Tshacher, W. & Dauwalder, J.-P.) 29–48 (World Scientific, 1999). · DOI
  32. Jordan, I. D., Sokół, P. A. & Park, I. M. Gated recurrent units viewed through the lens of continuous time dynamical systems. Front. Comput. Neurosci. 15, 678158 (2021). · DOI
  33. Beer, R. D. Codimension-2 parameter space structure of continuous-time recurrent neural networks. Biol. Cybern. 116, 501–515 (2022). · DOI
  34. Wang, X.-J. Theory of the multiregional neocortex: large-scale neural dynamics and distributed cognition. Annu. Rev. Neurosci. 45, 533–560 (2022). · DOI · Paper page
  35. O’Leary, T., Sutton, A. C. & Marder, E. Computational models in the age of large datasets. Curr. Opin. Neurobiol. 32, 87–94 (2015). · DOI
  36. Ostrow, M., Eisen, A. J., Kozachkov, L. & Fiete, I. R. Beyond geometry: comparing the temporal structure of computation in neural circuits with dynamical similarity analysis. In Proc. 37th Conference on Neural Information Processing Systems https://openreview.net/forum?id=7blSUMwe7R (OpenReview.net, 2023). · DOI
  37. Galgali, A. R., Sahani, M. & Mante, V. Residual dynamics resolves recurrent contributions to neural computation. Nat. Neurosci. 26, 326–338 (2023). · DOI
  38. Saxena, S., Russo, A. A., Cunningham, J. & Churchland, M. M. Motor cortex activity across movement speeds is predicted by network-level strategies for generating muscle activity. eLife 11, e67620 (2022). · DOI
  39. Arlt, C. et al. Cognitive experience alters cortical involvement in goal-directed navigation. eLife 11, e76051 (2022). · DOI
  40. Latimer, K. W. & Freedman, D. J. Low-dimensional encoding of decisions in parietal cortex reflects long-term training history. Nat. Commun. 14, 1010 (2023). · DOI
  41. Molano-Mazón, M. et al. Recurrent networks endowed with structural priors explain suboptimal animal behavior. Curr. Biol. 33, 622–638 (2023). · DOI
  42. Kingma, D. P., & Ba, J. Adam: a method for stochastic optimization. Proc. of the International Conference on Learning Representations (ICLR, 2015).
  43. Song, H. F., Yang, G. R. & Wang, X.-J. Reward-based training of recurrent neural networks for cognitive and value-based tasks. eLife 6, e21492 (2017). · DOI · Paper page
  44. Ward, J. H. Jr Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 58, 236–244 (1963). · DOI · Paper page

Cited by

See all 214

Metadata

Authors
Driscoll, L. N., Shenoy, K., & Sussillo, D.
Venue
Nature Neuroscience
Published
2024-07
Volume
27
Issue
7
Pages
1349-1363
ISSN
  • 1097-6256 (ISSN-L)
  • 1546-1726
OpenAlex
W4400462656
Cited by
214
References
18
FWCI
26.49
Type
journal article
Language
en
Open access
hybrid
Licence
Funders
Simons Foundation (543045, 1155867, 543049)

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Flexible multitask computation in recurrent networks utilizes shared dynamical motifs. https://app.cochat.ai/papers/10.1038/s41593-024-01668-6
The paper
Driscoll, L. N., Shenoy, K., & Sussillo, D. (2024). Flexible multitask computation in recurrent networks utilizes shared dynamical motifs. Nature Neuroscience. https://doi.org/10.1038/s41593-024-01668-6

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.