When Christopher Lynn set out to untangle the inner workings of individual neurons, he was bracing for chaos — but instead found elegant simplicity hiding inside the brain's most fundamental units. A Yale physicist has discovered that 90% of neuronal activity in mice operates via straightforward on-off switches, a finding that validates some of the earliest neuron models ever created, dating back to the 1940s.
This matters because for decades, neuroscientists have grappled with a paradox: the brain performs extraordinarily complex tasks — thinking, moving, speaking, remembering — yet individual neurons appear far less sophisticated than the systems they build. Lynn's new research, published in Nature Physics, suggests that neurons don't need to be complicated to support intelligence. They can be elegantly simple and still generate the rich computational power that emerges when billions of them work in concert.
Lynn, an assistant professor of physics at Yale's Faculty of Arts and Sciences and member of both the Quantitative Biology Institute and Wu Tsai Institute, spent years studying how neurons combine into networks. But for this project, he shifted his lens entirely, examining what happens inside a single cell. He built a computational model that parsed neuronal activity into three components: simple one-input-one-output interactions, complex interactions involving multiple inputs, and "latent noise," the inherent randomness that arises from synapses firing and the unpredictability of cellular chemistry.
To test his framework, Lynn analyzed brain data from mice and from C. elegans, a commonly studied worm. The results surprised even him. In mouse data, 90% of neuronal activity fell into the simple category — straightforward switch-like behavior. In worms, the figure was 60% to 70%. "I was expecting to see a roughly equal percentage for each of the three types," Lynn said. "I was so surprised. We have 100 billion nerve cells firing in our brain, and each one of them has 10,000 connections to other neurons. There was no reason to expect that a single neuron has such a simple description."
What makes this discovery particularly striking is its historical resonance. The earliest mathematical models of neurons date to 1943, when Warren S. McCulloch and Walter Pitts published their groundbreaking framework — and that 1940s model holds up remarkably well. Rather than rendering those early theories obsolete, modern neuroscience has simply confirmed what pioneers guessed through pure logic decades before we had the tools to measure neurons directly.
Lynn plans to extend this work, comparing his findings across other animal species to see whether neurons behave differently depending on the evolutionary complexity of the organism. The question becomes: as brains evolved greater capability, did individual neurons become more complicated, or did evolution instead refine how these simple units connect and communicate?
The answer could reshape how scientists think about the relationship between simplicity and complexity in biology — and why the most powerful systems sometimes emerge from the most straightforward building blocks.