Keywords at Cosyne

Analysis of more than 10 years’ worth of abstracts from the annual Cosyne meeting highlights research trends in systems and computational neuroscience.

The Simons Collaboration on the Global Brain launched in 2014 with the goal of studying populations of neurons at single-cell resolution. To explore research trends in areas that are important to the SCGB, we analyzed the frequency of selected keywords appearing in abstracts from the Computational and Systems Neuroscience (Cosyne) conference from 2010 to 2021.

One of the SCGB’s central goals has been to better understand neural coding and dynamics of populations of neurons at the single-cell level, a relatively novel endeavor when the program first launched. At the time, new technologies were just beginning to make it possible to simultaneously analyze the activity of hundreds of cells. That approach has since taken root — the frequency of keywords relating to neural dynamics has grown in recent years.

Understanding the structure of neural activity in large populations requires new analysis techniques, another major focus for the SCGB. Keywords related to the geometry of neural activity, including manifold, geometry, state space and dimensionality reduction, have also become more frequent in Cosyne abstracts, with a notable jump occurring in 2017.

Artificial neural networks have become a popular tool for modeling neural systems and probing potential computational mechanisms, as well as for analyzing behavior and other data.

Methods: In this project, we used data provided by the Cosyne conference organizers, including titles, abstracts, authors and other related information for all abstracts presented at annual Cosyne conferences from 2010 to 2021. A total of 4,400 abstracts were presented during this period, with a range of 340 to 420 abstracts per year. We used Microsoft Excel text matching formulas to identify studies that included different SCGB-related key terms in their titles and/or abstracts. We analyzed 18 key terms, which matched a number of studies ranging from 0 studies for one key term to 406 for another. For each key term, we plotted the number of studies presented by year to visualize changes in popularity. Out of the 18 key terms included in the analysis, we selected 11 to highlight in this story.