Document Open Access Logo

Contrastive Hebbian Learning for Multicomponent Liquids

Authors Yancheng Du , Cameron Chalk , Salvador Buse , Lulu Qian , Erik Winfree

PDF

File

Thumbnail PDF
PDF
LIPIcs.DNA.32.13.pdf
  • Filesize: 11.46 MB
  • 27 pages

Document Identifiers

Subject Classification

ACM Subject Classification
  • Hardware → Neural systems
  • Applied computing → Physics
  • Applied computing → Systems biology
Keywords
  • multicomponent liquid
  • liquid-liquid phase separation
  • continuous Hopfield network
  • contrastive Hebbian learning

Metrics

Abstract

Molecules with designed interactions can serve as substrates for information processing; demonstrated examples include algorithmic self-assembly of DNA tiles and DNA strand displacement reactions. These two well-established paradigms correspond to solid-phase-like behavior, where spatially structured molecular assemblies grow by crystalline attachment, and gas-phase-like behavior, where a dilute mixture of freely diffusing molecules is governed by mass-action kinetics of reactions. A third paradigm, liquid-liquid phase separation, is a fundamental phenomenon in physical and biological systems, giving rise to membraneless compartments that facilitate complex information processing. Recent studies have shown that multicomponent liquid mixtures described by lattice models are analogous to Boltzmann machines and can implement neural computation. Mean-field models have also suggested a connection to Hopfield networks and illustrated complex decision-making during condensation from a reservoir. As an alternative to previous approaches to train liquid interaction energies using backpropagation through a loss function, here we derive "local" learning rules based on contrastive Hebbian learning that may be more biologically and physically plausible. The trained systems demonstrate a range of computational tasks, including associative recall, linear and nonlinear input-output relations, pattern classification, and spatial phase separation. Our work suggests that the computational potential of liquid-phase molecular systems could be unlocked in real physical systems with local learning rules.

Cite As Get BibTex

Yancheng Du, Cameron Chalk, Salvador Buse, Lulu Qian, and Erik Winfree. Contrastive Hebbian Learning for Multicomponent Liquids. In 32nd International Conference on DNA Computing and Molecular Programming (DNA 32). Leibniz International Proceedings in Informatics (LIPIcs), Volume 387, pp. 13:1-13:27, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2026) https://doi.org/10.4230/LIPIcs.DNA.32.13

Author Details

Yancheng Du
  • California Institute of Technology, Pasadena, CA, USA
Cameron Chalk
  • California Institute of Technology, Pasadena, CA, USA
Salvador Buse
  • California Institute of Technology, Pasadena, CA, USA
Lulu Qian
  • California Institute of Technology, Pasadena, CA, USA
Erik Winfree
  • California Institute of Technology, Pasadena, CA, USA

Funding

  • Du, Yancheng: Schmidt Sciences Polymath Award to LQ.
  • Chalk, Cameron: NSF CCF/FET 2212546.
  • Buse, Salvador: Open Philanthropy Good Ventures Foundation, NSF CCF/FET Award 2212546.
  • Qian, Lulu: Schmidt Sciences Polymath Award, NSF CCF/FET 2212546.
  • Winfree, Erik: NSF CCF/FET 2212546.

Acknowledgements

The authors want to thank Inhoo Lee, Sandra Temgoua, Daichi Hayakawa, Arvind Murugan, Krishna Shrinivas, and others for helpful discussions.

Supplementary Materials

References

  1. David H Ackley, Geoffrey E Hinton, and Terrence J Sejnowski. A learning algorithm for Boltzmann machines. Cognitive Science, 9(1):147-169, 1985. URL: https://doi.org/10.1207/S15516709COG0901_7.
  2. Siddharth Agarwal, Dino Osmanovic, Mahdi Dizani, Melissa A Klocke, and Elisa Franco. Dynamic control of DNA condensation. Nature Communications, 15(1):1915, 2024. Google Scholar
  3. Blaise Agüera y Arcas, Jyrki Alakuijala, James Evans, Ben Laurie, Alexander Mordvintsev, Eyvind Niklasson, Ettore Randazzo, and Luca Versari. Computational life: How well-formed, self-replicating programs emerge from simple interaction. arXiv:2406.19108, 2024. Google Scholar
  4. Daniel J Amit, Hanoch Gutfreund, and Haim Sompolinsky. Storing infinite numbers of patterns in a spin-glass model of neural networks. Physical Review Letters, 55(14):1530, 1985. Google Scholar
  5. Daniel J Amit, Hanoch Gutfreund, and Haim Sompolinsky. Statistical mechanics of neural networks near saturation. Annals of Physics, 173(1):30-67, 1987. Google Scholar
  6. Minkyung Baek, Frank DiMaio, Ivan Anishchenko, Justas Dauparas, Sergey Ovchinnikov, Gyu Rie Lee, Jue Wang, Qian Cong, Lisa N Kinch, R Dustin Schaeffer, et al. Accurate prediction of protein structures and interactions using a three-track neural network. Science, 373(6557):871-876, 2021. Google Scholar
  7. Jack E Baldwin and Hans Krebs. The evolution of metabolic cycles. Nature, 291(5814):381-382, 1981. Google Scholar
  8. Salman F Banani, Hyun O Lee, Anthony A Hyman, and Michael K Rosen. Biomolecular condensates: organizers of cellular biochemistry. Nature Reviews Molecular Cell Biology, 18(5):285-298, 2017. Google Scholar
  9. Naama Barkai and Stan Leibler. Robustness in simple biochemical networks. Nature, 387(6636):913-917, 1997. Google Scholar
  10. MA Belyaev and AA Velichko. Classification of handwritten digits using the Hopfield network. In IOP Conference Series: Materials Science and Engineering, volume 862, page 052048. IOP Publishing, 2020. Google Scholar
  11. James Bradbury, Roy Frostig, Peter Hawkins, Matthew James Johnson, Yash Katariya, Chris Leary, Dougal Maclaurin, George Necula, Adam Paszke, Jake VanderPlas, Skye Wanderman-Milne, and Qiao Zhang. JAX: composable transformations of Python+NumPy programs, 2018. URL: http://github.com/jax-ml/jax.
  12. Rodrigo Braz Teixeira, Giorgio Carugno, Izaak Neri, and Pablo Sartori. Liquid Hopfield model: Retrieval and localization in multicomponent liquid mixtures. Proceedings of the National Academy of Sciences, 121(48):e2320504121, 2024. Google Scholar
  13. Jehoshua Bruck and Vwani P Roychowdhury. On the number of spurious memories in the Hopfield model (neural network). IEEE Transactions on Information Theory, 36(2):393-397, 1990. Google Scholar
  14. Miguel A Carreira-Perpinan and Geoffrey Hinton. On contrastive divergence learning. In International Workshop on Artificial Intelligence and Statistics, pages 33-40. PMLR, 2005. Google Scholar
  15. Aria S Chaderjian, Sam Wilken, and Omar A Saleh. Diverse, distinct, and densely packed DNA nanostar droplets. Proceedings of the National Academy of Sciences, 123(7):e2523462123, 2026. Google Scholar
  16. Cameron Chalk, Salvador Buse, Krishna Shrinivas, Arvind Murugan, and Erik Winfree. Learning and inference in a lattice model of multicomponent condensates. In 30th International Conference on DNA Computing and Molecular Programming (DNA 30), volume 314, pages 5:1-5:24, 2024. URL: https://doi.org/10.4230/LIPIcs.DNA.30.5.
  17. Fan Chen and William M Jacobs. Programmable phase behavior in fluids with designable interactions. The Journal of Chemical Physics, 158(21), 2023. Google Scholar
  18. Long Qing Chen and Jie Shen. Applications of semi-implicit Fourier-spectral method to phase field equations. Computer Physics Communications, 108(2-3):147-158, 1998. Google Scholar
  19. Kevin M Cherry and Lulu Qian. Scaling up molecular pattern recognition with DNA-based winner-take-all neural networks. Nature, 559(7714):370-376, 2018. Google Scholar
  20. Kevin M Cherry and Lulu Qian. Supervised learning in DNA neural networks. Nature, 645(8081):639-647, 2025. Google Scholar
  21. HE Cook. Brownian motion in spinodal decomposition. Acta Metallurgica, 18(3):297-306, 1970. Google Scholar
  22. David Doty, Jack H Lutz, Matthew J Patitz, Robert T Schweller, Scott M Summers, and Damien Woods. The tile assembly model is intrinsically universal. In 2012 IEEE 53rd Annual Symposium on Foundations of Computer Science, pages 302-310. IEEE, 2012. URL: https://doi.org/10.1109/FOCS.2012.76.
  23. Constantine G Evans and Erik Winfree. Physical principles for DNA tile self-assembly. Chemical Society Reviews, 46(12):3808-3829, 2017. Google Scholar
  24. Constantine Glen Evans, Jackson O’Brien, Erik Winfree, and Arvind Murugan. Pattern recognition in the nucleation kinetics of non-equilibrium self-assembly. Nature, 625(7995):500-507, 2024. Google Scholar
  25. Martin J Falk, Adam T Strupp, Benjamin Scellier, and Arvind Murugan. Temporal contrastive learning through implicit non-equilibrium memory. Nature Communications, 16(1):2163, 2025. Google Scholar
  26. Paul J Flory. Thermodynamics of high polymer solutions. The Journal of Chemical Physics, 10(1):51-61, 1942. Google Scholar
  27. Mark E Fornace, Jining Huang, Cody T Newman, Avinash Nanjundiah, Nicholas J Porubsky, Marshall B Pierce, and Niles A Pierce. NUPACK: Computational nucleic acid analysis and design. ACS Synthetic Biology, 15(4):1426-1441, 2026. Google Scholar
  28. Alireza Goudarzi, Matthew R Lakin, and Darko Stefanovic. DNA reservoir computing: a novel molecular computing approach. In International Workshop on DNA-Based Computers, pages 76-89. Springer, 2013. URL: https://doi.org/10.1007/978-3-319-01928-4_6.
  29. Geoffrey E Hinton. Deterministic Boltzmann learning performs steepest descent in weight-space. Neural Computation, 1(1):143-150, 1989. URL: https://doi.org/10.1162/NECO.1989.1.1.143.
  30. Pierre C Hohenberg and Bertrand I Halperin. Theory of dynamic critical phenomena. Reviews of Modern Physics, 49(3):435, 1977. Google Scholar
  31. John J Hopfield. Neural networks and physical systems with emergent collective computational abilities. Proceedings of the National Academy of Sciences, 79(8):2554-2558, 1982. Google Scholar
  32. John J Hopfield. Neurons with graded response have collective computational properties like those of two-state neurons. Proceedings of the National Academy of Sciences, 81(10):3088-3092, 1984. Google Scholar
  33. John J Hopfield and David W Tank. "Neural" computation of decisions in optimization problems. Biological Cybernetics, 52(3):141-152, 1985. Google Scholar
  34. Maurice L Huggins. Solutions of long chain compounds. The Journal of Chemical Physics, 9(5):440-440, 1941. Google Scholar
  35. Anthony A Hyman, Christoph A Weber, and Frank Jülicher. Liquid-liquid phase separation in biology. Annual Review of Cell and Developmental Biology, 30:39-58, 2014. Google Scholar
  36. William M Jacobs and Daan Frenkel. Predicting phase behavior in multicomponent mixtures. The Journal of Chemical Physics, 139(2):024108, 2013. Google Scholar
  37. William M Jacobs and Daan Frenkel. Phase transitions in biological systems with many components. Biophysical Journal, 112(4):683-691, 2017. Google Scholar
  38. Byoung-jin Jeon, Dan T Nguyen, and Omar A Saleh. Sequence-controlled adhesion and microemulsification in a two-phase system of DNA liquid droplets. The Journal of Physical Chemistry B, 124(40):8888-8895, 2020. Google Scholar
  39. Hawoong Jeong, Sean P Mason, A-L Barabási, and Zoltan N Oltvai. Lethality and centrality in protein networks. Nature, 411(6833):41-42, 2001. Google Scholar
  40. Patrick Kidger. On Neural Differential Equations. PhD thesis, University of Oxford, 2021. Google Scholar
  41. Shigeru Kondo and Takashi Miura. Reaction-diffusion model as a framework for understanding biological pattern formation. Science, 329(5999):1616-1620, 2010. Google Scholar
  42. Matthew R Lakin and Darko Stefanovic. Supervised learning in adaptive DNA strand displacement networks. ACS Synthetic Biology, 5(8):885-897, 2016. Google Scholar
  43. Inhoo Lee, Salvador Buse, and Erik Winfree. Differentiable programming of indexed chemical reaction networks and reaction-diffusion systems. In 31st International Conference on DNA Computing and Molecular Programming (DNA 31), pages 4:1-4:23, 2025. URL: https://doi.org/10.4230/LIPIcs.DNA.31.4.
  44. Timothy P Lillicrap, Adam Santoro, Luke Marris, Colin J Akerman, and Geoffrey Hinton. Backpropagation and the brain. Nature Reviews Neuroscience, 21(6):335-346, 2020. Google Scholar
  45. Fritz Lipmann. The ATP-phosphate cycle. In Current Topics in Cellular Regulation, volume 18, pages 301-311. Elsevier, 1981. Google Scholar
  46. Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. arXiv:1711.05101, 2017. Google Scholar
  47. Sheng Mao, Derek Kuldinow, Mikko P Haataja, and Andrej Košmrlj. Phase behavior and morphology of multicomponent liquid mixtures. Soft Matter, 15(6):1297-1311, 2019. Google Scholar
  48. Javier R Movellan. Contrastive Hebbian learning in the continuous Hopfield model. In Connectionist Models: Proceedings of the 1990 Summer School, pages 10-17. Morgan Kaufmann, 1991. Google Scholar
  49. Dan T Nguyen, Byoung-jin Jeon, Gabrielle R Abraham, and Omar A Saleh. Length-dependence and spatial structure of DNA partitioning into a DNA liquid. Langmuir, 35(46):14849-14854, 2019. Google Scholar
  50. Shu Okumura, Guillaume Gines, Nicolas Lobato-Dauzier, Alexandre Baccouche, Robin Deteix, Teruo Fujii, Yannick Rondelez, and Anthony J Genot. Nonlinear decision-making with enzymatic neural networks. Nature, 610(7932):496-501, 2022. Google Scholar
  51. Jacob Parres-Gold, Matthew Levine, Benjamin Emert, Andrew Stuart, and Michael B Elowitz. Contextual computation by competitive protein dimerization networks. Cell, 188(7):1984-2002, 2025. Google Scholar
  52. Lulu Qian, David Soloveichik, and Erik Winfree. Efficient Turing-universal computation with DNA polymers. In International Workshop on DNA-Based Computers, pages 123-140. Springer, 2010. URL: https://doi.org/10.1007/978-3-642-18305-8_12.
  53. Lulu Qian, Erik Winfree, and Jehoshua Bruck. Neural network computation with DNA strand displacement cascades. Nature, 475(7356):368-372, 2011. URL: https://doi.org/10.1038/NATURE10262.
  54. Yicheng Qiang, Chengjie Luo, and David Zwicker. Scaling laws for phase coexistence in multicomponent mixtures. Physical Review Research, 7(4):043008, 2025. Google Scholar
  55. Yusuke Sato, Tetsuro Sakamoto, and Masahiro Takinoue. Sequence-based engineering of dynamic functions of micrometer-sized DNA droplets. Science Advances, 6(23):eaba3471, 2020. Google Scholar
  56. Benjamin Scellier and Yoshua Bengio. Equilibrium propagation: Bridging the gap between energy-based models and backpropagation. Frontiers in Computational Neuroscience, 11:24, 2017. URL: https://doi.org/10.3389/FNCOM.2017.00024.
  57. Nadrian C. Seeman. Structural DNA Nanotechnology. Cambridge University Press, Cambridge, United Kingdom, 2016. Google Scholar
  58. Ludovica Serricchio, Dario Bocchi, Claudio Chilin, Raffaele Marino, Matteo Negri, Chiara Cammarota, and Federico Ricci-Tersenghi. Daydreaming Hopfield networks and their surprising effectiveness on correlated data. Neural Networks, 186:107216, 2025. URL: https://doi.org/10.1016/J.NEUNET.2025.107216.
  59. Krishna Shrinivas and Michael P Brenner. Multiphase coexistence capacity in complex fluids. bioRxiv:2022.10.19.512909, 2022. Google Scholar
  60. Krishna Shrinivas and Michael P Brenner. Phase separation in fluids with many interacting components. Proceedings of the National Academy of Sciences, 118(45):e2108551118, 2021. Google Scholar
  61. Krishna Shrinivas, Benjamin R Sabari, Eliot L Coffey, Isaac A Klein, Ann Boija, Alicia V Zamudio, Jurian Schuijers, Nancy M Hannett, Phillip A Sharp, Richard A Young, et al. Enhancer features that drive formation of transcriptional condensates. Molecular Cell, 75(3):549-561, 2019. Google Scholar
  62. Friedrich C Simmel, Bernard Yurke, and Hari R Singh. Principles and applications of nucleic acid strand displacement reactions. Chemical Reviews, 119(10):6326-6369, 2019. Google Scholar
  63. Karuna Skipper and Shelley FJ Wickham. DNA nanostars that self-assemble into core-shell condensate microdroplets. Nanoscale Horizons, 2026. Google Scholar
  64. David Soloveichik, Georg Seelig, and Erik Winfree. DNA as a universal substrate for chemical kinetics. Proceedings of the National Academy of Sciences, 107(12):5393-5398, 2010. Google Scholar
  65. Tianqi Song and Lulu Qian. Heat-rechargeable computation in DNA logic circuits and neural networks. Nature, 646(8084):315-322, 2025. Google Scholar
  66. Ulrich Stelzl, Uwe Worm, Maciej Lalowski, Christian Haenig, Felix H Brembeck, Heike Goehler, Martin Stroedicke, Martina Zenkner, Anke Schoenherr, Susanne Koeppen, et al. A human protein-protein interaction network: a resource for annotating the proteome. Cell, 122(6):957-968, 2005. Google Scholar
  67. Menachem Stern and Arvind Murugan. Learning without neurons in physical systems. Annual Review of Condensed Matter Physics, 14(1):417-441, 2023. Google Scholar
  68. Jonathan Stricker, Tobias Falzone, and Margaret L Gardel. Mechanics of the F-actin cytoskeleton. Journal of Biomechanics, 43(1):9-14, 2010. Google Scholar
  69. Diana A Tanase, Dino Osmanović, Roger Rubio-Sánchez, Layla Malouf, Elisa Franco, and Lorenzo Di Michele. Internal phase separation in synthetic DNA condensates. Advanced Science, 12(41):e06275, 2025. Google Scholar
  70. Rodrigo Braz Teixeira, Izaak Neri, and Pablo Sartori. Metastable phase separation and information retrieval in multicomponent mixtures. arXiv:2509.10705, 2025. Google Scholar
  71. Gašper Tkačik and William Bialek. Information processing in living systems. Annual Review of Condensed Matter Physics, 7(1):89-117, 2016. Google Scholar
  72. Hirotake Udono, Jing Gong, Yusuke Sato, and Masahiro Takinoue. DNA droplets: intelligent, dynamic fluid. Advanced Biology, 7(3):2200180, 2023. Google Scholar
  73. Jan van den Berg. The most general framework of continuous Hopfield neural networks. In Proceedings of International Workshop on Neural Networks for Identification, Control, Robotics and Signal/Image Processing, pages 92-100. IEEE, 1996. Google Scholar
  74. Jan van den Berg. Generalized Hopfield networks for constrained optimization. Technical Report EUR-FEW-CS-98-04, Erasmus School of Economics, Erasmus University Rotterdam, 1998. URL: http://hdl.handle.net/1765/761.
  75. Christian Waltermann and Edda Klipp. Information theory based approaches to cellular signaling. Biochimica et Biophysica Acta (BBA)-General Subjects, 1810(10):924-932, 2011. Google Scholar
  76. Erik Winfree. Algorithmic self-assembly of DNA: Theoretical motivations and 2D assembly experiments. Journal of Biomolecular Structure and Dynamics, 17(sup1):263-270, 2000. Google Scholar
  77. Erik Winfree, Xiaoping Yang, and Nadrian C Seeman. Universal computation via self-assembly of DNA: Some theory and experiments. In DNA Based Computers II, volume 44 of DIMACS: Series in Discrete Mathematics and Theoretical Computer Science, pages 191-213. American Mathematical Society, 1998. Google Scholar
  78. Damien Woods, David Doty, Cameron Myhrvold, Joy Hui, Felix Zhou, Peng Yin, and Erik Winfree. Diverse and robust molecular algorithms using reprogrammable DNA self-assembly. Nature, 567(7748):366-372, 2019. URL: https://doi.org/10.1038/S41586-019-1014-9.
  79. Xiaohui Xie and H Sebastian Seung. Equivalence of backpropagation and contrastive Hebbian learning in a layered network. Neural Computation, 15(2):441-454, 2003. URL: https://doi.org/10.1162/089976603762552988.
  80. Wataru Yahiro, Nathanael Aubert-Kato, and Masami Hagiya. A reservoir computing approach for molecular computing. In Artificial Life Conference Proceedings, pages 31-38. MIT Press, 2018. URL: https://doi.org/10.1162/ISAL_A_00013.
  81. Aidan Zentner, Ethan V Halingstad, Cameron Chalk, Michael P Brenner, Arvind Murugan, Erik Winfree, and Krishna Shrinivas. Information processing driven by multicomponent surface condensates. arXiv:2509.08100, 2025. Google Scholar
  82. David Yu Zhang and Georg Seelig. Dynamic DNA nanotechnology using strand-displacement reactions. Nature Chemistry, 3(2):103-113, 2011. Google Scholar
  83. David Zwicker and Liedewij Laan. Evolved interactions stabilize many coexisting phases in multicomponent liquids. Proceedings of the National Academy of Sciences, 119(28):e2201250119, 2022. Google Scholar
Any Issues?
X

Feedback on the Current Page

CAPTCHA

Thanks for your feedback!

Feedback submitted to Dagstuhl Publishing

Could not send message

Please try again later or send an E-mail
© 2023-2026 Schloss Dagstuhl – LZI GmbH Schloss Dagstuhl – LZI GmbH About DROPS Imprint Privacy Contact