#AcademicRunPlaylist - 6/29/24

It was a looong day, but at least I was able to listen to lots of talks for my #AcademicRunPlaylist!

First was a great talk by Aniket Rege on learning nested representations for a variety of models at UW-Madison Computer Sciences https://www.youtube.com/watch?v=IbfdvzPwZTg

Next was a nice talk by Tianqi Chen on bringing large models to consumer devices at the Stanford University Department of Computer Science https://www.youtube.com/watch?v=InoNMvjs_vo

Next was a fantastic conversation with Joshua Vogelstein on connectomes, using statistics with neural data, the problems with AI doomers, and more on the Brain Inspired podcast. Highly recommend https://www.youtube.com/watch?v=V_UkhWeP49w

Next was a compelling talk by Glenn Magerman on identifying relevant markets using firm embeddings at INET Oxford. I still wonder how sensitive this approach is to different settings, but it could definitely help cut through the noise around antitrust market definition https://www.youtube.com/watch?v=UVWfw1OsH2g

Next was an excellent talk by Liran Morav on understanding economics through social relationships (with tons of Ron Burt shout outs!) at the Cambridge Society for Economic Pluralism https://www.youtube.com/watch?v=OziBKCYNwrc

Next was an intriguing talk by Giannis Daras on learning consistent diffusion models at UW Madison https://www.youtube.com/watch?v=qf5AZ62nO6E

Next was an interesting talk by Tim Dettmers on using k-bit quantization to reduce training resources for large models at Stanford https://www.youtube.com/watch?v=EsMcVkTXZrk

Next was an engaging talk by Yuchen Zeng on getting multimodal LLMs to perform text-to-image in context learning at UW Madison https://www.youtube.com/watch?v=GNLkneJWnEE

Next was a wide-ranging symposium on domestic workers in contemporary Latin American culture with Rachel Randall, Geoffrey Kantaris, Sonia Roncador, and María Julia Rossi at the University of London School of Advanced Study https://www.youtube.com/watch?v=ZJUxcUqFGc8

Next was an informative talk by Haotian Liu on building customizable multimodal models at UW Madison https://www.youtube.com/watch?v=To8Gzo3IPlc

Next was a fascinating talk by Roman Roth on Italy in the second century BCE at the University of London School of Advanced Study https://www.youtube.com/watch?v=YJf4kvWHnU0

Next was a thought-provoking talk by Fanghui Liu on examining if overfitted adversarially trained networks can generalize from an approximation viewpoint at UW Madison https://www.youtube.com/watch?v=6mqGmNRo7Po

Next was a timely talk by David Weisstanner on income decline and the political backlash against advanced welfare states (😬) at INET Oxford https://www.youtube.com/watch?v=D1rxqhWfLcI

Last was a good talk by Harit Vishwakarma on the promises and pitfalls of threshold-based auto-labeling at UW Madison https://www.youtube.com/watch?v=WYUKyBy2yRs