1932
0
  1. Home
  2. A-Z Publications
  3. Annual Review of Condensed Matter Physics
  4. Volume 14, 2023
  5. Article

Annual Review of Condensed Matter Physics

Volume 14, 2023

Learning Without Neurons in Physical Systems

Abstract

Learning is traditionally studied in biological or computational systems. The power of learning frameworks in solving hard inverse problems provides an appealing case for the development of physical learning in which physical systems adopt desirable properties on their own without computational design. It was recently realized that large classes of physical systems can physically learn through local learning rules, autonomously adapting their parameters in response to observed examples of use. We review recent work in the emerging field of physical learning, describing theoretical and experimental advances in areas ranging from molecular self-assembly to flow networks and mechanical materials. Physical learning machines provide multiple practical advantages over computer designed ones, in particular by not requiring an accurate model of the system, and their ability to autonomously adapt to changing needs over time. As theoretical constructs, physical learning machines afford a novel perspective on how physical constraints modify abstract learning theory.

Keyword(s): inverse design, learning theory, machine learning, metamaterials, molecular computing, physical learning, self-assembly
Loading

Article metrics loading...

/content/journals/10.1146/annurev-conmatphys-040821-113439
2023-03-10
2024-04-26
Loading full text...

Full text loading...

/deliver/fulltext/conmatphys/14/1/annurev-conmatphys-040821-113439.html?itemId=/content/journals/10.1146/annurev-conmatphys-040821-113439&mimeType=html&fmt=ahah

Literature Cited

  1. 1.
    De Vito E, Rosasco L, Caponnetto A, De Giovannini U, Odone F. 2005. J. Mach. Learn. Res. 6:883–904
  2. 2.
    Murugan A, Zeravcic Z, Brenner MP, Leibler S. 2015. PNAS 112:54–59
  3. 3.
    Stern M, Pinson MB, Murugan A. 2020. Phys. Rev. X 10:3031044
  4. 4.
    Pashine N, Hexner D, Liu AJ, Nagel SR. 2019. Sci. Adv. 5:12eaax4215
  5. 5.
    Stern M, Arinze C, Perez L, Palmer SE, Murugan A. 2020. PNAS 117:2614843–50
  6. 6.
    Stern M, Hexner D, Rocks JW, Liu AJ. 2021. Phys. Rev. X 11:2021045
  7. 7.
    Pedretti G, Milo V, Ambrogio S, Carboni R, Bianchi S et al. 2017. Sci. Rep. 7:5288
  8. 8.
    Hebb DO. 2005. The Organization of Behavior: A Neuropsychological Theory London: Psychology
  9. 9.
    Micali G, Endres RG. 2016. Curr. Opin. Microbiol. 30:8–15
  10. 10.
    Bull MS, Prakash VN, Prakash M. 2021. arXiv:2107.02934
  11. 11.
    Marbach S, Ziethen N, Bastin L, Baeuerle F, Alim K. 2021. bioRxiv: 2021.12.29.474405
  12. 12.
    Ristroph L, Ristroph G, Morozova S, Bergou AJ, Chang S et al. 2013. J. R. Soc. Interface 10:8520130237
  13. 13.
    Peleg O, Peters JM, Salcedo MK, Mahadevan L. 2018. Nat. Phys. 14:121193–98
  14. 14.
    Braitenberg V. 1986. Vehicles: Experiments in Synthetic Psychology Cambridge, MA: MIT Press
  15. 15.
    Kanai R, Komura Y, Shipp S, Friston K. 2015. Philos. Trans. R. Soc. B: Biol. Sci. 370:166820140169
  16. 16.
    Tlusty T, Libchaber A, Eckmann JP. 2017. Phys. Rev. X 7:2021037
  17. 17.
    Husain K, Murugan A. 2020. Mol. Biol. Evol. 37:102865–74
  18. 18.
    Wang Y, Chen M, Zhou F, Ma E 2002. Nature 419:6910912–15
  19. 19.
    Karbasian H, Tekkaya AE. 2010. J. Mater. Proc. Technol. 210:152103–18
  20. 20.
    Baird DG, Collias DI. 2014. Polymer Processing: Principles and Design Hoboken, NJ: Wiley
  21. 21.
    Keim NC, Paulsen JD, Zeravcic Z, Sastry S, Nagel SR. 2019. Rev. Mod. Phys. 91:3035002
  22. 22.
    Lillicrap TP, Santoro A, Marris L, Akerman CJ, Hinton G. 2020. Nat. Rev. Neurosci. 21:6335–46
  23. 23.
    Nemenman I. 2005. Neural Comput. 17:92006–33
  24. 24.
    Caporale N, Dan Y 2008. Annu. Rev. Neurosci. 31:25–46
  25. 25.
    Bengio Y, Lee DH, Bornschein J, Mesnard T, Lin Z. 2015. arXiv:1502.04156
  26. 26.
    Scellier B. 2021. A deep learning theory for neural networks grounded in physics PhD Thesis, Univ. Montr., Montreal, Quebec Can: arXiv:2103.09985
  27. 27.
    Dillavou S, Stern M, Liu AJ, Durian DJ. 2022. Phys. Rev. Appl. 18:014040
  28. 28.
    Wycoff JF, Dillavou S, Stern M, Liu AJ, Durian DJ. 2022. J. Chem. Phys. 156:14144903
  29. 29.
    Burr GW, Shelby RM, Sebastian A, Kim S, Kim S et al. 2017. Adv. Phys.: X 2:89–124
  30. 30.
    Marković D, Mizrahi A, Querlioz D, Grollier J. 2020. Nat. Rev. Phys. 2:9499–510
  31. 31.
    Piccinini G. 2015. Physical Computation: A Mechanistic Account Oxford, UK: Oxford Univ. Press
  32. 32.
    Adleman LM. 1994. Science 266:51871021–24
  33. 33.
    Soloveichik D, Cook M, Winfree E, Bruck J. 2008. Nat. Comput. 7:4615–33
  34. 34.
    Tanaka G, Yamane T, Héroux JB, Nakane R, Kanazawa N et al. 2019. Neural Netw. 115:100–23
  35. 35.
    Jaeger H, Haas H. 2004. Science 304:566778–80
  36. 36.
    Silverberg JL, Evans AA, McLeod L, Hayward RC, Hull T et al. 2014. Science 345:6197647–50
  37. 37.
    Bertoldi K, Vitelli V, Christensen J, Van Hecke M. 2017. Nat. Rev. Mater. 2:1117066
  38. 38.
    Rocks JW, Pashine N, Bischofberger I, Goodrich CP, Liu AJ, Nagel SR. 2017. PNAS 114:102520–25
  39. 39.
    Hexner D, Pashine N, Liu AJ, Nagel SR. 2020. Phys. Rev. Res. 2:4043231
  40. 40.
    Hexner D, Liu AJ, Nagel SR. 2020. PNAS 117:5031690–95
  41. 41.
    Hagh VF, Nagel SR, Liu AJ, Manning ML, Corwin EI. 2022. PNAS 119:19e2117622119
  42. 42.
    Zhong W, Schwab DJ, Murugan A. 2017. J. Stat. Phys. 167:3806–26
  43. 43.
    Evans CG, O'Brien J, Winfree E, Murugan A 2022. arxiv:2207.06399
  44. 44.
    Wei B, Dai M, Yin P. 2012. Nature 485:623–26
  45. 45.
    Kramar M, Alim K 2021. PNAS 118:10e2007815118
  46. 46.
    Lin X, Rivenson Y, Yardimci NT, Veli M, Luo Y et al. 2018. Science 361:64061004–8
  47. 47.
    Wright LG, Onodera T, Stein MM, Wang T, Schachter DT et al. 2022. Nature 601:7894549–55
  48. 48.
    Dudte LH, Vouga E, Tachi T, Mahadevan L. 2016. Nat. Mater. 15:5583–88
  49. 49.
    Stern M, Jayaram V, Murugan A. 2018. Nat. Commun. 9:4303
  50. 50.
    Stern M, Pinson MB, Murugan A. 2017. Phys. Rev. X 7:4041070
  51. 50a.
    Arinze C, Stern M, Nagel SR, Murugan A. 2022. arXiv:2206.08886
  52. 51.
    Bray D. 1995. Nature 376:6538307–12
  53. 52.
    Winfree E. 1998. Algorithmic self-assembly of DNA. PhD Thesis, Calif. Inst. Technol., Pasadena, CA
  54. 53.
    Cherry KM, Qian L. 2018. Nature 559:7714370–76
  55. 54.
    Zeravcic Z, Manoharan VN, Brenner MP. 2017. Rev. Mod. Phys. 89:3031001
  56. 55.
    Biffi S, Cerbino R, Bomboi F, Paraboschi EM, Asselta R et al. 2013. PNAS 110:3915633–37
  57. 56.
    Needleman D, Dogic Z. 2017. Nat. Rev. Mater. 2:917048
  58. 57.
    Schaus TE, Woo S, Xuan F, Chen X, Yin P 2017. Nat. Commun. 8:696
  59. 58.
    Fredriksson S, Gullberg M, Jarvius J, Olsson C, Pietras K et al. 2002. Nat. Biotechnol. 20:5473–77
  60. 59.
    Lee JH, Lee SH, Baek C, Chun H, Ryu Jh et al. 2017. Biosystems 158:1–9
  61. 60.
    Baek C, Lee SW, Lee BJ, Kwak DH, Zhang BT. 2019. Molecules 24:71409
  62. 61.
    Baum EB. 1995. Science 268:5210583–85
  63. 62.
    Mills AP, Yurke B, Platzman PM. 1999. Biosystems 52:175–80
  64. 63.
    Zhong W, Schwab DJ, Murugan A. 2017. J. Stat. Phys. 167:3806–26
  65. 64.
    Hopfield JJ. 1982. PNAS 79:82554–58
  66. 65.
    Jacobs WM. 2021. Phys. Rev. Lett. 126:25258101
  67. 66.
    Shrinivas K, Brenner MP. 2021. PNAS 118:45e2108551118
  68. 67.
    Lee JB, Peng S, Yang D, Roh YH, Funabashi H et al. 2012. Nat. Nanotechnol. 7:12816–20
  69. 68.
    Mohammed AM, Sulc P, Zenk J, Schulman R. 2017. Nat. Nanotechnol. 12:4312–16
  70. 69.
    Schulman R, Yurke B, Winfree E. 2012. PNAS 109:176405–10
  71. 70.
    Gandhi N, Ashkenasy G, Tannenbaum E. 2007. J. Theor. Biol. 249:58–66
  72. 71.
    Lakin MR, Stefanovic D. 2016. ACS Synthet. Biol. 5:8885–97
  73. 72.
    Poole W, Ortiz-Muñoz A, Behera A, Jones NS, Ouldridge TE et al. 2017. DNA Computing and Molecular Programming. DNA 23. Lecture Notes in Computer Science, Vol. 10467 R Brijder, L Qian 210–31. Cham, Switz: Springer
  74. 73.
    Tayar AM, Hagan MF, Dogic Z. 2021. PNAS 118:30e2102873118
  75. 74.
    Saha S, Nagy TL, Weiner OD. 2018. Philos. Trans. R. Soc. B Biol. Sci. 373:174720170145
  76. 75.
    Majumdar S, Foucard LC, Levine AJ, Gardel ML. 2018. Soft Matter 14:112052–58
  77. 76.
    Li S, Batra R, Brown D, Chang HD, Ranganathan N et al. 2019. Nature 567:7748361–65
  78. 77.
    Cichos F, Gustavsson K, Mehlig B, Volpe G. 2020. Nat. Mach. Intel. 2:294–103
  79. 78.
    Falk MJ, Alizadehyazdi V, Jaeger H, Murugan A. 2021. Phys. Rev. Res. 3:3033291
  80. 79.
    Rubenstein M, Cornejo A, Nagpal R. 2014. Science 345:6198795–99
  81. 80.
    Werfel J, Petersen K, Nagpal R. 2014. Science 343:6172754–58
  82. 81.
    Banerjee S, Gardel ML, Schwarz US. 2020. Annu. Rev. Condens. Matter Phys. 11:421–39
  83. 82.
    Tero A, Takagi S, Saigusa T, Ito K, Bebber DP et al. 2010. Science 327:5964439–42
  84. 83.
    Katifori E, Szöllősi GJ, Magnasco MO. 2010. Phys. Rev. Lett. 104:4048704
  85. 84.
    Ronellenfitsch H, Katifori E. 2016. Phys. Rev. Lett. 117:13138301
  86. 85.
    Anisetti VR, Scellier B, Schwarz JM. 2022. arXiv:2203.12098
  87. 86.
    Kim KH, Gaba S, Wheeler D, Cruz-Albrecht JM, Hussain T et al. 2012. Nano Lett. 12:389–95
  88. 87.
    Grollier J, Querlioz D, Camsari K, Everschor-Sitte K, Fukami S, Stiles MD. 2020. Nat. Electron. 3:7360–70
  89. 88.
    Rosenthal E, Greshnikov S, Soudry D, Kvatinsky S. 2016. 2016 IEEE International Symposium on Circuits and Systems (ISCAS)1394–97. Piscataway, NJ: IEEE
  90. 89.
    Serrano-Gotarredona T, Masquelier T, Prodromakis T, Indiveri G, Linares-Barranco B. 2013. Front. Neurosci. 7:2
  91. 90.
    Kendall J, Pantone R, Manickavasagam K, Bengio Y, Scellier B. 2020. arXiv:2006.01981
  92. 91.
    Martin E, Ernoult M, Laydevant J, Li S, Querlioz D et al. 2021. iScience 24:3102222
  93. 92.
    Sui X, Wu Q, Liu J, Chen Q, Gu G 2020. IEEE Access 8:70773–83
  94. 93.
    Rajendran B, Sebastian A, Schmuker M, Srinivasa N, Eleftheriou E. 2019. IEEE Signal Proc. Mag. 36:697–110
  95. 94.
    Agrawal DK, Jiang R, Reinhart S, Mohammed AM, Jorgenson TD, Schulman R. 2017. ACS Nano 11:109770–79
  96. 95.
    Pashine N. 2021. Phys. Rev. Mater. 5:6065607
  97. 96.
    Cheng Z, Ríos C, Pernice WH, Wright CD, Bhaskaran H. 2017. Sci. Adv. 3:9e1700160
  98. 97.
    Bisker G, England JL. 2018. PNAS 115:45E10531–38
  99. 98.
    Fisher RA. 1936. Ann. Eugen. 7:2179–88
  100. 99.
    Movellan JR 1991. Connectionist Models: Proceedings of the 1990 Summer School DS Touretzky, JL Elman, TJ Sejnowski, GE Hinton 10–17. San Mateo, CA: Morgan Kaufmann Publ.
  101. 100.
    Xie S, Seung HS. 2003. Neural Comput. 15:2441–54
  102. 101.
    Scellier B, Bengio Y. 2017. Front. Comput. Neurosci. 11:24
  103. 102.
    Stern M, Dillavou S, Miskin MZ, Durian DJ, Liu AJ. 2022. Phys. Rev. Res. 4:2L022037
  104. 103.
    Lopez-Pastor V, Marquardt F. 2021. arXiv:2103.04992
  105. 104.
    Scellier B, Goyal A, Binas J, Mesnard T, Bengio Y. 2018. arXiv:1808.04873
  106. 105.
    Kim S, Du C, Sheridan P, Ma W, Choi S, Lu WD. 2015. Nano Lett. 15:32203–11
  107. 106.
    Serb A, Bill J, Khiat A, Berdan R, Legenstein R, Prodromakis T. 2016. Nat. Commun. 7:12611
  108. 107.
    Anisetti VR, Kandala A, Scellier B, Schwarz J. 2022. arXiv:2208.08862
  109. 108.
    Bernstein J, Wang YX, Azizzadenesheli K, Anandkumar A. 2018. Proceed. Mach. Learn. Res. 80:560–69
  110. 109.
    Murugan A, Zou J, Brenner MP. 2015. Nat. Commun. 6:6203
  111. 110.
    Ventriglia F, Di Maio V. 2002. Biosystems 67:1–3287–94
  112. 111.
    Keim NC, Nagel SR. 2011. Phys. Rev. Lett. 107:010603
  113. 112.
    Bottou L. 2003. Advanced Lectures on Machine Learning146–68. Berlin: Springer
  114. 113.
    McKnight G, Henry C 2005. Smart Structures and Materials 2005: Active Materials: Behavior and Mechanics, Vol. 5761 WA Armstrong 119–26. Bellingham, WA: SPIE
  115. 114.
    Lagoudas DC. 2008. Shape Memory Alloys: Modeling and Engineering Applications New York: Springer
  116. 115.
    Henke M, Gerlach G. 2014. Microsyst. Technol. 20:4–5599–606
  117. 116.
    Zhou Y, Duque CM, Santangelo CD, Hayward RC. 2019. Adv. Funct. Mater. 29:481905273
  118. 117.
    Sorscher B, Mel G, Ganguli S, Ocko S. 2019. Adv. Neural Inf. Process. Syst. 32:10003–13
  119. 118.
    Hu M, Strachan JP, Li Z, Grafals EM, Davila N et al. 2016. 2016 53nd ACM/EDAC/IEEE Design Automation Conference (DAC), Austin, Texas, June 5–9 Hewlett Packard Labs HPE-2016-23. https://www.labs.hpe.com/techreports/2016/HPE-2016-23.pdf
  120. 119.
    Wang Z, Li C, Song W, Rao M, Belkin D et al. 2019. Nat. Electron. 2:3115–24
  121. 120.
    Amari Si, Murata N, Müller KR, Finke M, Yang H. 1995. Adv. Neural Inf. Process. Syst. 8:145–51
  122. 121.
    Bhaumik H, Hexner D. 2022. Phys. Rev. Res 44L042044
  123. 122.
    Zucker RS, Regehr WG. 2002. Annu. Rev. Physiol. 64:355–405
  124. 123.
    Marom S. 2010. Prog. Neurobiol. 90:16–28
  125. 124.
    Ernoult M, Grollier J, Querlioz D, Bengio Y, Scellier B. 2019. Adv. Neural Inform. Proc. Syst. 32:7081–91
  126. 125.
    Bartunov S, Santoro A, Richards BA, Marris L, Hinton GE, Lillicrap T. 2018. arXiv:1807.04587
  127. 126.
    Landauer R. 1961. IBM J. Res. Dev. 5:3183–91
  128. 127.
    Fruchart M, Hanai R, Littlewood PB, Vitelli V. 2021. Nature 592:7854363–69
  129. 128.
    Seliger P, Young SC, Tsimring LS. 2002. Phys. Rev. E 65:4041906
  130. 129.
    French RM. 1999. Trends Cogn. Sci. 3:4128–35
  131. 130.
    Bahri Y, Kadmon J, Pennington J, Schoenholz SS, Sohl-Dickstein J, Ganguli S. 2020. Annu. Rev. Condens. Matter Phys. 11:501–28
  132. 131.
    Rocks JW, Mehta P. 2022. Phys. Rev. Res. 4:013201
  133. 132.
    Hertz J, Krogh A, Palmer RG. 2018. Introduction to the Theory of Neural Computation Boca Raton, FL: CRC
  134. 133.
    Fink TM, Ball RC. 2001. Phys. Rev. Lett. 87:19198103
  135. 134.
    Hinton GE, Osindero S, Teh YW. 2006. Neural Comput. 18:71527–54
  136. 135.
    Coulais C, Teomy E, De Reus K, Shokef Y, Van Hecke M. 2016. Nature 535:7613529–32
  137. 136.
    Pinson MB, Witten TA. 2016. J. Phys. Condens. Matter 28:49495102
  138. 137.
    Mézard M, Parisi G, Virasoro MA. 1987. Spin Glass Theory and Beyond: An Introduction to the Replica Method and Its Applications, Vol. 9 Singapore: World Sci.
  139. 138.
    Wagner GP, Altenberg L. 1996. Evolution 50:3967–76
  140. 139.
    Johnson ME, Hummer G. 2011. PNAS 108:2603–8
  141. 140.
    Huntley MH, Murugan A, Brenner MP. 2016. PNAS 113:215841–46
  142. 141.
    Su CJ, Murugan A, Linton JM, Yeluri A, Bois J et al. 2022. Cell Syst 13:5408–25.e12
  143. 142.
    Alon U. 2007. Nat. Rev. Genet. 8:6450–61
  144. 143.
    Modes CD, Magnasco MO, Katifori E. 2016. Phys. Rev. X 6:3031009
  145. 144.
    Kashtan N, Alon U. 2005. PNAS 102:3913773–78
  146. 145.
    Hemery M, Rivoire O. 2015. Phys. Rev. E 91:4042704
  147. 146.
    Bloom JD, Labthavikul ST, Otey CR, Arnold FH. 2006. PNAS 103:155869–74
  148. 147.
    Murugan A, Jaeger HM. 2019. MRS Bull. 44:296–105
  149. 148.
    Falk MJ, Wu J, Matthews A, Sachdeva V, Pashine N et al. 2022. arXiv:2211.02270
  150. 149.
    Pinson MB, Stern M, Carruthers Ferrero A, Witten TA, Chen E, Murugan A 2017. Nat. Commun. 8:15477
  151. 150.
    Yan L, Ravasio R, Brito C, Wyart M. 2018. Biophys. J. 114:122787–98
  152. 151.
    Recanatesi S, Farrell M, Lajoie G, Deneve S, Rigotti M, Shea-Brown E. 2021. Nat. Commun. 12:1417
  153. 152.
    Kaneko K, Tsuda I. 2003. Chaos: Interdiscip. J. Nonlinear Sci. 13:3926–36
  154. 153.
    Transtrum MK, Machta BB, Brown KS, Daniels BC, Myers CR, Sethna JP. 2015. J. Chem. Phys. 143:07B201_1
/content/journals/10.1146/annurev-conmatphys-040821-113439
Loading
Learning Without Neurons in Physical Systems
Annual Review of Condensed Matter Physics 14, 417 (2023); https://doi.org/10.1146/annurev-conmatphys-040821-113439
/content/journals/10.1146/annurev-conmatphys-040821-113439
/content/journals/10.1146/annurev-conmatphys-040821-113439
Loading

Data & Media loading...

  • Article Type: Review Article

Most Cited Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error