Sublime
An inspiration engine for ideas

Doing Matrix Calculus can be messy, specially when we need higher order derivatives.
Writing them out using Tensor Diagrams makes even the Hessian Chain Rule relatively simple:
FreeGaussian: Guidance-free Controllable 3D Gaussian Splats with Flow Derivatives
Contributions (cited):
• We propose FreeGaussian, a novel annotations guidance-free Gaussian Splatting method for
controllable scene reconstruction, which automatically explores interactable scene structures... See more
MrNeRFx.com
This is a great high-level introduction to Hamiltonian Monte Carlo methods that I found very accessible.
It has great visual explanations of a variety of relevant concepts. https://t.co/0YI8OYHDNR
Exceedingly crispy multi-view 3D captures reconstructed beautifully (try it in browser below).
Instead of the usual "deform one master model over time" approach (which breaks w/ fast motion) this FreeTimeGS paper spawns ephemeral gaussian particles as needed.
They live briefly, move as... See more
Bilawal Sidhux.com

I've finally solved steepest descent on Finsler-structured (matrix) manifolds more generally. This generalizes work by me, @jxbz, and @Jianlin_S on Muon, Orthogonal Muon, & Stiefel Muon.
---
The general solution turned out to be much simpler than I thought. And it should generalize to any... See more

Natural gradient with momentum on submanifolds:
In Riemannian normal coordinate systems, metric tensor becomes identity but computationally intractable.
We introduce generalized normal coordinates https://t.co/D82DD38hQz


