#AcademicRunPlaylist - 9/13/24

A selfie of me on a sunny day in front of a pond, with tree branches hanging overhead. I'm a bald, middle-aged, white man with a red beard flecked with white. I'm wearing black sunglasses and a red running shirt

It was another hectic day, but I was able to fit in a shorter run along with some talks for my #AcademicRunPlaylist!

First was the 4th workshop of adversarial machine learning on computer vision at #CVPR2024. I highly recommend the whole event, with standout talks by Zico Kolter (adversarial attacks on aligned language models, talk of the year candidate) and Ludwig Schmidt (data-centric view on reliable generalization) https://www.youtube.com/watch?v=U3SiUQvZ5LM

Next was the first day of the National Bureau of Economic Research wage dynamics in the 21st century symposium. I particularly liked the talk by Ben Etheridge on the impact of labor demand shocks when occupational labor supplies are heterogeneous https://www.youtube.com/watch?v=r8KiiWjPlD4

Next was a pair of talks by Jake Sunshine (using smart devices to detect cardiopulmonary emergencies) and Roni Sengupta (neural rendering and its application in healthcare) at CVPR https://www.youtube.com/watch?v=bvBHsiyAhiA

Next was an incredible talk by Phillip Isola on why large models are all converging to similar representations at MIT Brain and Cognitive Sciences. I absolutely love this work, and while I think the convergence here is on the world as represented by the Internet rather than the entire world as Isola hypothesizes, this approach should fundamentally reshape how companies and researchers value training more models and why phenomena transfer across models (see Kolter's talk above, for example). Highly recommend https://www.youtube.com/watch?v=V7AyriUcXZQ

Next was an interesting talk by Mark Ho on abstraction in humans and machines at CVPR https://youtu.be/LooLbLs3O_Y?si=iGO7cR9WKiDQo9m_&t=5210

Last was the learning from procedural videos and language workshop at CVPR. I especially liked the talk by Dima Damen on methods for going from fine-grained to global understanding of these videos https://www.youtube.com/watch?v=cyhPkJuEviI