How can we design neural networks in a principled way? New work from DeWeese Lab Berkeley AI Research explores a theoretically-motivated paradigm for doing just that for dense networks! Read about it here: bair.berkeley.edu/blog/2022/08/2…

just updated deweeselab.com! check it out for a view of new research and new lab members.

In 2020, Michael Nielsen & I began a 2-month project to write: "how would we fund science?" 2 years & 40,000 words later, it's become: "how can the culture & institutions of science actually change, and ultimately become self-improving?" Our answer: scienceplusplus.org/metascience/in…

Lab member Dhruva just revamped the NTK wikipedia entry to make it more accessible to technical non-experts. cool gif inside -- check it out! en.wikipedia.org/wiki/Neural_ta…

Happy to see Jamie's research about the stepwise nature of SSL on the BAIR blog: bairblog.github.io/2023/07/10/ste… If you're interested in learning more, stop by and say hi to us at ICML next week!

Found a kindred spirit at #icml2023 #feelthelearn