Divisive normalization. Many of my efforts have focused on a neural computation called divisive normalization, whereby the activity of a neuron is divided by the summed activity of other neurons. As a PhD student I refined the model, I proposed a biophysical substrate for it based on synaptic inhibition, and I provided the first direct support for it in single neurons of the primate visual cortex. In my laboratory, we then extended the model to single neurons in earlier stages of the visual system and to entire populations of neurons in visual cortex, thus showing that a single equation can capture the activity of a large population of cortical neurons. These results and others led me to propose that normalization is a canonical neural computation, one that takes place in multiple neural systems to serve multiple functions. 

Cortical inhibition. Related efforts concern the roles of cortical inhibition. As a postdoc, I showed that inhibition in visual cortex has the same sensory selectivity as excitation. In my laboratory, we studied the role of inhibition with population recordings paired with pharmacology, and with intracellular recordings paired with optogenetics. The results indicate that normalization is not due to inhibition, but rather to changes in mutual excitation. We then provided the first measurements of inhibition in individual cortical neurons during wakefulness, revealing that in the awake cortex, inhibition dominates responses.

Cortical connectivity. Further efforts center on cortical connectivity and on its impact on neuronal populations. Using optical imaging and parallel electrodes, we showed that visual stimuli elicit travelling waves of activity in visual cortex, and that such waves depend on brain state. We also found a law relating the selectivity of individual neurons to their location in visual cortex. We then discovered that synaptic connectivity makes some cortical neurons respond in tune with the nearby population an others independently of it (“choristers” vs. “soloists”), and that this difference can be explained by differences in excitatory connectivity and results in a simplified statistical description of cortical activity. 

Sensory adaptation. A longstanding area of interest are the phenomena of perceptual and neural adaptation that follow prolonged presentation of sensory stimuli. As a postdoc, I discovered a cellular basis for adaptation in the cerebral cortex: a steady hyperpolarization. Later work in my laboratory revealed the effect of adaptation on large neuronal populations, and showed how adaptation acts as a homeostatic mechanism that equalizes response levels across time. Further work revealed how adaptation cascades from one brain region to the subsequent one. 

Nonsensory modulation. Neurons in cortex respond with high variability when one repeats a sensory stimulus many times. We showed that this variability grows markedly between the thalamus and the visual cortex, and that its effects are amplified inside individual cortical neurons. Later work showed that the main sources of variability in visual cortex are shared across neurons, and that much of variability in visual cortex can be explained by behavioral factors such as locomotion and navigation. These factors affect visual cortex as much as visual stimuli.