{"id":360692,"date":"2026-04-02T19:03:10","date_gmt":"2026-04-02T19:03:10","guid":{"rendered":"https:\/\/www.newsbeep.com\/nz\/360692\/"},"modified":"2026-04-02T19:03:10","modified_gmt":"2026-04-02T19:03:10","slug":"digital-sphinx-raises-questions-about-connectome-models","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/nz\/360692\/","title":{"rendered":"\u2018Digital sphinx\u2019 raises questions about connectome models"},"content":{"rendered":"<p>A neural network based on a nematode worm\u2019s connectome can puppeteer a digital fruit fly\u2019s body, a new <a href=\"https:\/\/doi.org\/10.64898\/2026.03.20.713233\" target=\"_blank\" rel=\"nofollow noopener\">preprint<\/a> shows. The work comes just two weeks after Eon Systems, a neurotechnology company based in San Francisco, announced that it had \u201c<a href=\"https:\/\/eon.systems\/updates\/weve-uploaded-a-fruit-fly\" target=\"_blank\" rel=\"nofollow noopener\">uploaded<\/a>\u201d a fly brain and released a video of that brain controlling a biomechanical fly model in a virtual world.<\/p>\n<p>\u201cWe need to be really careful in interpreting this kind of work,\u201d says <a href=\"https:\/\/www.bingbrunton.com\/people\" target=\"_blank\" rel=\"nofollow noopener\">Bing Wen Brunton<\/a>, professor of biology at the University of Washington, who posted the new preprint on bioRxiv in response to Eon Systems\u2019 announcement. Working with her team, Brunton strung a biophysical model of a Drosophila body to a simulation of the Caenorhabditis elegans connectome and trained that \u201cdigital sphinx,\u201d a term coined in the preprint,\u00a0 to walk using deep reinforcement learning\u2014all with a \u201cbrain\u201d that wasn\u2019t a fly brain at all.\u00a0<\/p>\n<p>Brunton\u2019s work points to an important control for other researchers looking to combine deep learning and <a href=\"https:\/\/doi.org\/10.1038\/s41586-024-07939-3\" target=\"_blank\" rel=\"nofollow noopener\">connectomics<\/a> to simulate fly behavior, says <a href=\"https:\/\/www.cshl.edu\/research\/faculty-staff\/benjamin-cowley\/\" target=\"_blank\" rel=\"nofollow noopener\">Benjamin Cowley<\/a>, assistant professor at Cold Spring Harbor Laboratory, who was not involved in the preprint. It enables them to ask, \u201cIf I just created a randomly connected connectome, could it also do the same behaviors?\u201d he says.<\/p>\n<p><p class=\"first-letter:text-2xl first-letter:float-left first-letter:text-red first-letter:pr-2 pb-0\">\n    T<\/p>\n<p>he problem with connectome models is that they do not capture the biophysical properties of neurons or the pools of neurotransmitters that modulate neural communication, and they exist without a body, says <a href=\"https:\/\/www.janelia.org\/people\/srinivas-turaga\" target=\"_blank\" rel=\"nofollow noopener\">Srinivas Turaga<\/a>, group leader at the Howard Hughes Medical Institute\u2019s Janelia Research Campus, who was not involved in the work.\u00a0<\/p>\n<p>To circumvent these shortcomings, researchers are starting to use <a href=\"https:\/\/arxiv.org\/pdf\/2602.17997\" target=\"_blank\" rel=\"nofollow noopener\">deep reinforcement learning<\/a> to relate the connectomes to behavior. But using these techniques to model biological processes also has pitfalls, as illustrated by the model Brunton and her team created.\u00a0<\/p>\n<p>Deep reinforcement learning is a process of optimization, Brunton says, and biological systems don\u2019t always work optimally. This approach can work \u201creally well\u201d in capturing fly behavior, she says, even when the model isn\u2019t biologically realistic.\u00a0<\/p>\n<p class=\"font-light italic text-2sm p-0 lg:text-3md\">\n            For lack of a better term, there\u2019s so much BS out there.        <\/p>\n<p> \u2014 <\/p>\n<p>                 \u2014<br \/>\n                Bing Wen Brunton <\/p>\n<p><a href=\"https:\/\/eon.systems\/\" target=\"_blank\" rel=\"nofollow noopener\">Eon<\/a> assembled its fly from three previously published datasets: a <a href=\"https:\/\/doi.org\/10.1038\/s41592-024-02497-y\" target=\"_blank\" rel=\"nofollow noopener\">biophysical model of the fly body<\/a> called NeuroMechFly, a fly brain connectome, and part of the fly visual system. And the company also used deep reinforcement learning to stitch the pieces together, training the networks to emulate a walking fly. But the Eon video quickly received <a href=\"https:\/\/www.theverge.com\/ai-artificial-intelligence\/894587\/fly-brain-computer-upload\" target=\"_blank\" rel=\"nofollow noopener\">pushback<\/a>, and Turaga says that any random network connected to the NeuroMechFly model in this way might generate a walking fly. Because Eon hasn\u2019t published the specifics of how they built their fly, it\u2019s unclear if the model is any more accurate than a random network, Turaga says.\u00a0<\/p>\n<p><a href=\"https:\/\/eon.systems\/#team\" target=\"_blank\" rel=\"nofollow noopener\">Philip Shiu<\/a>, head of engineering at Eon Systems, doesn\u2019t entirely disagree. \u201cI think it\u2019s fair to say that this is not a full blown copy of a fly. My personal preference might be to say maybe we ought to call this a digital twin or an embodied model,\u201d he says. \u201cObviously, part of the intention of the company was to say: This is something that\u2019s really exciting and cool and not science fiction as it has been in the past.\u201d<\/p>\n<p>Even a \u201csmall network of 300 neurons\u201d contains enough information for deep learning to extract patterns that drive realistic behaviors, Cowley adds. \u201cThere\u2019s enough randomness in this network that you can map it to fly legs and make them move in a reasonable way.\u201d\u00a0<\/p>\n<p><p class=\"first-letter:text-2xl first-letter:float-left first-letter:text-red first-letter:pr-2 pb-0\">\n    B<\/p>\n<p>runton was familiar with Eon\u2019s work before it was released, she says; early last year, the company approached her and her colleague, <a href=\"https:\/\/nbio.uw.edu\/people\/entry\/ec0e1de9\/\" target=\"_blank\" rel=\"nofollow noopener\">John Tuthill<\/a>, professor of neurobiology and biophysics at the University of Washington and an investigator on the preprint, in hopes of collaborating, though that never came to fruition. Brunton says she also saw Eon present a poster on its virtual fly at last year\u2019s Society for Neuroscience conference.\u00a0<\/p>\n","protected":false},"excerpt":{"rendered":"A neural network based on a nematode worm\u2019s connectome can puppeteer a digital fruit fly\u2019s body, a new&hellip;\n","protected":false},"author":2,"featured_media":360693,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[111,139,69,145],"class_list":{"0":"post-360692","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-technology","8":"tag-new-zealand","9":"tag-newzealand","10":"tag-nz","11":"tag-technology"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts\/360692","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/comments?post=360692"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts\/360692\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/media\/360693"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/media?parent=360692"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/categories?post=360692"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/tags?post=360692"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}