nextfx
Re:Genie 3: A new frontier for world models
on Tuesday August 05, 2025 @07:05PM (#44805613)
Another case of moving the goalposts until you score a goal.
Inside Google Willow
Re:Numpyro: Probabilistic programming with NumPy powered by Jax
on Sunday November 17, 2024 @04:36PM (#42167535)
I'm not an expert on MCMC for BNNs by any means, but even with small networks I think it's a little tricky to get right. If my memory serves me right, this paper[1] focuses on small networks and goes over the issues and how to get around them.
Re:Numpyro: Probabilistic programming with NumPy powered by Jax
on Sunday November 17, 2024 @08:41AM (#42164146)
MCMC is AFAIU overly prohibitive for neural networks. If you are interested in incorporating uncertainty awareness and improving calibration of your neural net classifiers in a somewhat scalable manner, I think linearized Laplace is a good place to look. I'd suggest you look at `laplace-torch` [1], it's an easy way of doing that.
Re:FUSE-T is a kext-less implementation of FUSE for macOS that uses NFSv4
on Tuesday December 26, 2023 @03:35PM (#38775776)
I've been using mksquashfs to make full-system backups lately, but had to resort to using unsquashfs for recovering files. This combined with squashfuse might come in handy.
Re:I Implemented Nyströmformer
on Monday August 22, 2022 @04:47PM (#32556988)
I assume the name is referring to the Nyström low-rank matrix approximation.
Re:Expose server behind NAT with Tor
on Friday January 14, 2022 @02:24PM (#29939085)
Is there a way of doing NAT hole-punching with Wireguard?
Ask HN: Should I major or minor in Machine Learning?
I recently started my Masters degree in university and I have been trying to decide on what to major in. I find that, in my school, the Machine Learning major has a very limited scope of subjects and there are majors with broader subjects that also interest me. However I worry that if I want to apply for an “ML-heavy” position in the future, then majoring would be better.