Giving AAAI2024 tutorial recent advance in physics-informed machine learning with Grant Rotskoff tmw
Initial slide can be accessed through our website: sites.google.com/view/aaai-2024… See you all tmw!

Come join us for an informal meetup at #NeurIPS23 on Thursday 6-8pm at the conference venue (location marked in purple on the map near Hall C) to talk about #AI + Science & Math with a focus on neural operators and theorem proving.
Check out our latest library updates to make it…

#Neurips2023 come to our poster 'Geometry-Informed Neural Operator for Large-Scale 3D PDEs' on Tue at 11am, #500. We learn the solution operator of 3d RANS simulation with varying geometries.
Prof. Anima Anandkumar
Nikola Kovachki
Jean Kossaifi
Kamyar Azizzadenesheli

What if we set GPT-4 free in Minecraft? ⛏️

I’m excited to announce Voyager, the first lifelong learning agent that plays Minecraft purely in-context. Voyager continuously improves itself by writing, refining, committing, and retrieving *code* from a skill library.

GPT-4 unlocks…

Good morning from Caltech! 😊

Using ML to solve PDEs is a hot topic. With PDEBench we propose a comprehensive and extensible benchmark that includes multiple datasets, (pre-trained) baseline models such as the FNO and U-Net, as well as JAX-based numerical solvers for realistic PDEs. arxiv.org/abs/2210.07182

ArXiv arxiv.org/abs/2207.05209: Fourier neural operator for PDEs on arbitrary geometries. Learns to map the irregular input grid to a regular latent space grid, where the FFT is applied. 10^5 times faster than numerical solvers and 2x as accurate as existing ML-based PDE solvers.

New video! Partial Differential Equation (PDE) Overview

youtube.com/watch?v=pvrIag…

Super excited to be deriving and solving PDEs in this series!!! Next up: heat equation in 1D, 2D, ND; separation of variables; wave equation; Navier-Stokes equations, and more!!!

Another great example of #AIforScience We design a variant of Fourier neural operator #FNO that achieves for the first time via/metal layer contour simulation at 1nm^2/pixel resolution with any tile size in #lithography Zongyi Li NVIDIA

FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators
abs: arxiv.org/abs/2202.11214

Proud to see our #Fourier #neural operator #FNO listed among highlights in #math #Computer #science by Quanta Magazine It is transforming scientific simulations in many areas with orders of magnitude speedup. quantamagazine.org/the-year-in-ma… Zongyi Li @kazizzad Caltech NVIDIA AI

Check out our new arxiv preprint on efficient token mixers for transformers via Fourier neural operators arxiv.org/abs/2111.13587

Fourier meets attention but more efficiently O(nlogn). With soft thresholding it allows dynamic scaling to high/continuous resolution vision.

Remind me the good old days doing research on orbifolds 🤓

Physics-Informed Neural Operator for Learning Partial Differential Equations
deepai.org/publication/ph…
by Zongyi Li et al. including Prof. Anima Anandkumar
#NeuralNetwork #PartialDifferentialEquation

I am looking for PhD students to join my effort on ML for trustworthy AI fall 2022. Consider applying to our PhD program in the CS department of JHU JHU Computer Science Johns Hopkins Engineering. We have a great program and active multidisciplinary research. More info: cs.jhu.edu/phd-graduate-a…

The reading list for pilot run of 10721: Philosophical Foundations of Machine Intelligence has been finalized and presenter notes from groups to date are up on the course GitHub for anyone interested in reading along:

github.com/acmi-lab/cmu-1…

Hair care

Our paper on learning a PDE solution optimally is now on arxiv: arxiv.org/abs/2110.06897
an introduction can be find here: youtube.com/watch?v=Z5rtT5…