#AcademicRunPlaylist - 6/18/24

After being burnt to a crisp yesterday I spent a bit more time inside today, but later in the day I was able to go up to Portsmouth for a bit and explore while listening to talks for my #AcademicRunPlaylist!

First was a great talk by Michael Mahoney on bringing random matrix theory to bear on modern machine learning at DIMACS https://www.youtube.com/watch?v=7LwXiyx1QHo

Next was a short talk by Louis Kang on distinguishing concepts while building examples in hippocampal and artificial neural networks at the Simons Institute for the Theory of Computing https://www.youtube.com/watch?v=3oArAF9M4Lg

Next was a thought-provoking talk by Annelise Riles on lessons from socio-legal studies for a fractured world at the University of Oxford Faculty of Law https://www.youtube.com/watch?v=L4rwL7-xwWU

Next was a fascinating talk by Gily Ginosar on action-specific spatial coding in the medial entorhinal cortex of flying bats (!) at the Simons Institute https://www.youtube.com/watch?v=zLjUKSf7OiQ

Next was an interesting talk by Zhichao Wang on signal propagation and feature learning in neural networks at DIMACS https://www.youtube.com/watch?v=ZeC7jhJTAto

Next was a compelling talk by Kaylene Stocking on predicting brain activity in a driving task using activations in a deep learning network at the Simons Institute https://www.youtube.com/watch?v=fYr8_9gZ-DA

Next was an intriguing talk by Jerker Moodysson on innovation, institutional change, and policy innovation at Heidelberg University https://www.youtube.com/watch?v=rY4kc-zzr9Y

Last was an amazing talk by Alex Williams on the impact of noise on neural computation at the Simons Institute. After some motivations about the essential role noise has played in many recent ML advances, Williams uses neural data from monkeys (😬) on reaching tasks to illustrate the role of noise, how to model it more effectively in neural systems, and the importance of noise correlation structure. Highly recommend https://www.youtube.com/watch?v=tcY0MZQGqWE