An international collaboration of over 50 neuroscientists and researchers is pioneering a new approach to understanding the brain: the world’s first crowd-sourced and designed neuroscience study. The goal is to shed light on predictive processing, the brain’s ability to anticipate future events based on past experiences. The experiments began this week at the Allen Institute. 

Our brains are constantly predicting the future, whether it’s expecting a bear after hearing a rustle in the forest or processing everyday visual information. 

“Predictive coding is the idea that most of your brain areas might not be really encoding what you’re seeing, but rather, might be most interested in what’s different about what you’re seeing compared to what you expected,” explains Colleen Gillon, Ph.D., a researcher at Imperial College London who played a leading role in the collaboration. 

Using the Allen Institute’s state-of-the-art OpenScope platform, a set of tools to record signals from large populations of neurons during behavior, researchers have designed experiments to test key brain functions related to predictive processing. The study aims to uncover the trade-offs that the brain makes in different contexts when predicting what comes next. For example, imagine you’re sitting on a train. If the train is moving, you naturally expect the scenery outside to flow continuously; and many brain areas are working together to predict what happens next in the long, uninterrupted sequence. But if the train is stopped and you only see a single person walking by, your brain taps into a more specialized process to anticipate that individual’s movement and uses a different set of computations. 

A New Model for Neuroscience Research

The Allen Institute, supported by the Cognitive Computational Neuroscience conference, organized this first-of-its-kind international effort. A group of more than 50 scientists conducted extensive online collaboration, including more than 1,900 proposals and comments, to design the crowd-sourced study. This unprecedented level of cooperation will help the global scientific community better understand this important brain adaptation that is critical to survival. 

“Despite nearly 3,000 scientific publications on the topic, current theories have often been criticized as too vague or difficult to validate experimentally. This disconnect has hampered progress in both understanding brain function and applying these insights to advances in machine learning and artificial intelligence,” noted Jérôme Lecoq, Ph.D., associate investigator at the Allen Institute for Neural Dynamics.  

This online, collaborative model represents a potential paradigm shift in how neuroscience research is conducted. By pooling expertise from scientists around the globe and engaging in transparent discussion and decision-making, the project successfully integrates theory with experiments. The findings are expected to not only deepen our understanding of predictive processing but also inspire future collaborative efforts in neuroscience.