This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| Cosyne | |
|---|---|
| Name | Cosyne |
| Discipline | Neuroscience; Computational Neuroscience |
| Frequency | Annual |
| Established | 2004 |
| Organizers | Society for Neuroscience; Massachusetts Institute of Technology; independent academic organizers |
| Typical location | United States; primarily San Diego, Salt Lake City, San Francisco |
| Attendance | ~600–1000 |
Cosyne is an annual meeting focused on computational approaches to systems and sensory neuroscience that brings together researchers from laboratories across academia, industry, and government. The meeting emphasizes short presentations, interactive poster sessions, and direct discussion among scientists from institutions such as University of California, Berkeley, Massachusetts Institute of Technology, University College London, Max Planck Society, and Cold Spring Harbor Laboratory. Cosyne has become a hub connecting researchers working on topics related to the visual cortex, hippocampus, prefrontal cortex, deep learning, and statistical methods rooted in the traditions of Bayesian probability and information theory.
Cosyne was founded in 2004 by a small group of investigators seeking a focused venue complementary to larger gatherings like Society for Neuroscience and NeurIPS. Early meetings featured contributors from labs led by William Bialek, Terrence Sejnowski, Dmitri Chklovskii, and Laurence Abbott, and attracted participants associated with institutions such as Harvard University, Princeton University, California Institute of Technology, University of Pennsylvania, and Columbia University. Over time, the meeting expanded to include investigators from European centers including University of Oxford, École Normale Supérieure, Max Planck Institute for Biological Cybernetics, and Institut Pasteur, and collaborators from technology companies like Google DeepMind, Facebook AI Research, and Microsoft Research. Cosyne’s format evolved to prioritize short talks and extended poster sessions, fostering interactions similar to workshops at Gatsby Computational Neuroscience Unit and symposia at Cold Spring Harbor Laboratory.
Cosyne is organized each year by an ad hoc committee of senior and early-career organizers drawn from universities and research institutes such as Johns Hopkins University, Stanford University, University of Washington, University of California, San Diego, and New York University. The program typically includes plenary talks, short oral presentations selected from submitted abstracts, and extensive poster sessions modeled after formats used at EMBO workshops and Society for Neuroscience meetings. The meeting venue varies among cities including Salt Lake City, San Diego Convention Center, and conference centers near San Francisco International Airport, enabling participation from nearby labs at UCSF and Stanford Medicine. Submission procedures emphasize concise abstracts and figures, and peer reviewers often include faculty from Brown University, Yale University, Duke University, and the University of Chicago.
Cosyne’s scientific program spans a range of topics from experimental systems neuroscience to theoretical and computational methods. Recurring themes include sensory processing in systems such as the retina, primary visual cortex, and auditory cortex; navigation and memory in the hippocampus and entorhinal cortex; decision-making in the prefrontal cortex and basal ganglia; and methods like neural decoding, latent variable modeling, and reinforcement learning influenced by work at DeepMind and in the Reinforcement Learning community. Presentations frequently draw on analytic techniques from Bayesian statistics, maximum likelihood estimation, sparse coding, principal component analysis, and algorithms related to convolutional neural networks developed at Geoffrey Hinton-associated groups and industrial labs such as OpenAI.
Attendees include principal investigators, postdoctoral researchers, graduate students, and engineers from institutions and companies including NIH, Wellcome Trust, European Research Council, Allen Institute for Brain Science, MIT-IBM Watson AI Lab, and startup labs spun out of Columbia University. The community fosters cross-fertilization between experimentalists and theorists from groups led by figures such as Earl Miller, Mikhail Lebedev, Karel Svoboda, and Jakob Hohwy, and encourages participation from underrepresented regions through travel awards sponsored by organizations like Simons Foundation and Gordon and Betty Moore Foundation. Social events, unconference sessions, and software-sharing workshops often feature tools developed at Matlab, Python Software Foundation-affiliated projects, TensorFlow, and PyTorch user groups.
Cosyne sponsors awards and recognitions aimed at early-career scientists and outstanding presentations. Typical honors mirror scholarly recognition seen in prizes like the Turing Award and fellowships such as those from the National Science Foundation and Wellcome Trust. Specific meeting awards have included best poster and best talk distinctions judged by panels with members from European Molecular Biology Laboratory, Salk Institute, Columbia University, and Princeton Neuroscience Institute. Receipt of a Cosyne presentation is often cited on CVs alongside grants from agencies such as NIH and fellowships like the EMBO Young Investigator Programme.
Cosyne has had a sustained influence on the development of computational neuroscience by accelerating collaborations that led to publications in journals such as Nature Neuroscience, Neuron, Journal of Neuroscience, PLoS Computational Biology, and eLife. Techniques showcased at the meeting have propagated into industrial research at Google Research and clinical initiatives at Massachusetts General Hospital and University College London Hospitals. The meeting’s emphasis on reproducible code and open data contributed to practices later adopted by projects like the Allen Brain Atlas and protocols advocated by the BRAIN Initiative.
Notable contributions first presented at Cosyne include early demonstrations of population decoding methods that appeared in Nature Neuroscience and algorithms bridging recurrent neural network models with electrophysiology results cited alongside work from Laurence Abbott and Karel Svoboda. Seminal posters and talks have anticipated influential papers from labs such as Surya Ganguli, Ken Miller, Kaitlin S. O’Connell, Loren Frank, and Tom Mrsic-Flogel. Presentations often precede formal publication in venues like Proceedings of the National Academy of Sciences, Science, and Nature Communications, and are discussed in follow-up workshops at Gatsby Unit and NeurIPS satellite events.
Category:Computational neuroscience conferences