Adam and I discuss the possibility there are cost functions optimized in the brain, where and how it might work, Marvin Minsky’s Society of Mind, and lots more.
Blake and I discuss his recent work exploring one way credit assignment could occur in the brain, using different regions within a single cortical neuron to encode the necessary signal.
Josh and I talk about all the ways supervised machine learning can be used in neuroscience research, and we walk through how a variety of machine learning algorithms perform decoding on a few neural data sets.
In this episode, Dan Yamins and I talk about how he uses hierarchical convolutional neural networks to model the ventral visual stream, finding that the units in his model correspond to neurons in progressive layers of the brain. We also delve into the AI agents he develops that learn how to play through intrinsic motivation. Click the episode to get the show notes.
Ryan and I go deep (pun intended) on convolutional neural networks and how he uses them to solve problems in medical risk factor discovery and improve genome sequencing, and more. Click the episode for the show notes
