@jiaxinwen22 This is roughly like saying linear algebra cannot explain general relativity at all
If you've arrived at that conclusion, you've horrifically conflated the scope and purpose of the tool
Baby #2 is due any day now, and I am too wired to do the intelligent thing (i.e., sleep).
So please join me, the wonderful @osumray, and @haharrington on a journey through some fun machine learning & representation theory.
1/11
@mleingang@AlexKontorovich Latex is great for assignments in proof-based math classes. If you have to actually write out a computation however, LaTeX can be a nightmare. Especially if it involves typesetting matrices...
@irenecourtin @erinkburrell It was the first gen! It's possible that the new one is better, I haven't looked into it. My thought was that since it was a software issue, if the issues I have were not fixed in an update, then the new one might have them as well. Maybe not 🤷
It’s here! You can now apply for student debt cancellation from the Biden administration—and it only takes a couple of minutes:
https://t.co/TvwENHxRBb
I’ll be honest: I’ve stopped listening to the “VR is dead” folks altogether. Most of these takes fall into one of a handful of categories:
- I tried it years ago and it wasn’t good
- It can’t replace real-life
- I can still see the pixels
- Screens in your face are bad
etc. etc.
@linkedin_user01 Empathy is key. Turn their disdain into confusion when you agree about how poorly it was taught in primary schools, and how you wish you had known what it was really about earlier. Have them walk away questioning their understanding of what math *is*
MYTH: We don't have a solution to nuclear's "waste problem"
REALITY: Nuclear waste isn't a problem. In fact, it’s the best solution we have to meeting our energy needs while protecting the natural environment!
Here's what you need to know:
@FoldMani@vineettiruvadi The beauty of the dynamics of physical systems is that locally, they are linear. You just need to find the right "local" framing, and your linear tools are perfectly valid :)
DeepMath 2021 talks are now online! Check out all the fantastic speakers we've had over the past four years:
2021: https://t.co/Gy1HImKtQl
2020: https://t.co/TtO0OghcSs
2019: https://t.co/LNMbIrzoM9
2018: https://t.co/FYkhcEx9qi
@vineettiruvadi Not all phenomena that would naturally admit a neural representation have differentiable structure. Certainly the aspect of structure preservation is however of fundamental importance, whatever that structure may be.
Just capped off a semester-long informal discussion group among Hopkins PhD students on #GeometricDeepLearning with a fantastic Q&A session with @PetarV_93. Thanks to everyone who participated, both today and over all of our previous sessions! #MachineLearning
This was a incredible privilege to attend, and it is an even greater honor to be able to count myself among such a brilliant, driven, and kind group of individuals from all over the globe. Anyone interested in the future of ML would do well to keep an eye on this group!
60 students are attending our 1st in-person #machinelearningtheory summer school this week. Organizer @BorisHanin, assistant prof of operations research & financial engineering, aims to promote "a common language" to connect the next gen. of researchers in the field.
Finally out "GraFT: Graph Filtered Temporal Dictionary Learning for Functional Neural Imaging" w/ @gmishne @gehirnkarte @Natecermak & @SchillerJackie
GraFT uses a Graph based on pixel correlations to flexibly demix calcium imaging data across anatomical scales.