Dr. Leopoldo Petreanu is a Principal Investigator at the Champalimaud Center for the Unknown where he leads a Cortical Circuits Lab.
Dr. Petreanu is interested in the principles of inter- and intra-cortical connectivity. He levels an appropriately multi-pronged attack on this fundamental problem. For example, to better understand cortical feedback, he studied visual cortex and the functional organization therein. While when looking at the long-range circuits, he chose to look at the sensory and motor cortices. Aside from effortlessly switching between preparations and nailing complex techniques, his lab also develops novel behavioral tasks.
In this episode, we discuss
–> what Dr. Petreanu learnt from Karel Svoboda
–> why not having quantitative skills early on is not a handicap
–> which argument for the unity of cortical computation he finds the most compelling
1) What is the most successful theory in Neuroscience today?
Predictive coding. It certainly makes sense of a lot of findings in the cortical field.
Backpropagation in Deep Nets and how they can be implemented in the brain.
2) What is a recent piece of data you are most excited about?
Complete wiring diagrams of several circuits. But I think we need more novel theories to bring clarity to the volumes of data we can amass these days with contemporary methods.
As I mention in the episode, I thought the diagrams Dr. Petreanu used to relate his work to other findings in the field and provide a broader theoretical context for the audience were exceptionally efficient. Here are two examples courtesy of Dr. Petreanu.