Curious (more so confused) grad student. Simulation @Roblox, PhD @CASUSscience, prev, @GeorgiaTech & @Livermore_Lab. I ∈ {Science ∩ ML}.
Santa Clara
Joined July 2015
- Tweets226
- Following756
- Followers154
- Likes2.1K
Karan Shah retweeted
🚀 In a preprint with @ReKarantNetwork on electron dynamics simulations, we've accelerated state-of-the-art time-dependent #DFT using #NeuralOperators.
#MachineLearning #compchem #ICML2024
@helmholtz_ai @HZDR_Dresden @CASUSscience @icmlconf
doi.org/10.48550/arXiv.2407.…
Karan Shah retweeted
I am pleased to announce that our new research group website is now online.
Check it out: sites.google.com/view/mlmd/
Keep up to date on our research activities in #MachineLearning for #Materials and stay tuned for more news.
#compchem #compphys #theochem #condmat #AI
Karan Shah retweeted
Great new work by @ReKarantNetwork @AttilaCangi @opsimaths et al @UCMChemistry @UCMerced_NatSci @HZDR_Dresden @CASUSscience @helmholtz_ai- 'Inverting the Kohn–Sham equations with #physics-informed #machinelearning'-iopscience.iop.org/article/1… #PINNs #compchem #DFT #materials #HPC #AI
Karan Shah retweeted
Density functional theory is powerful, but finding the right exchange-correlation (XC) potential is tricky. Our new paper with @ucmerced explores how #AI can solve this fundamental problem. #compchem @HZDR_Dresden @CASUSscience @helmholtz_ai @IOPPublishing
tinyurl.com/ys4apd4c
Karan Shah retweeted
In our preprint we explore how #NeuralNetworks are used for inverting the Kohn-Sham equations of #DensityFunctionalTheory.
#materialsscience #compchem #machinelearning
@HZDR_Dresden @CASUSscience @ucmerced
@helmholtz_ai @ReKarantNetwork @arxiv
arxiv.org/abs/2312.15301
Karan Shah retweeted
Excited to share our new book chapter in the @SpringerNature book series Challenges and Advances in #compchem and #physics: #MachineLearning for Static and Dynamic #ElectronicStructureTheory.
#CondMat #MatSci
@HZDR_Dresden @CASUSscience
doi.org/10.1007/978-3-031-37…
Karan Shah retweeted
We are excited to announce that the Data Science Education Community of Practice (DSECOP) project has been featured in an article published in APS News on April 13, 2023. See the article here: buff.ly/3UReROc
Karan Shah retweeted
Thanks to #DPG for organizing a great #SKM2023 Spring Meeting . My group's research at @HZDR_Dresden and @CASUSscience on #MachineLearning and #materialsscience is featured in the following talks and poster. It is not too late to catch us for a discussion at @tudresden_de.
Karan Shah retweeted
Gifts to expect from your nerd boyfriend 🌻
@ReKarantNetwork
"A smiling elephant standing on a sunflower"
@nocontextmemes @OpenAI @Dalle #dalle2
Karan Shah retweeted
Imagine a nice, simple, low-entropy cup of coffee.
Pour some milk and you'll see beautiful & complex structures arising, before the cup reaches a high-entropy homogenous status, back to simple.
Interesting things happen while entropy increases!
Here a funny experiment:
1/3
Karan Shah retweeted
There's a budding tension in deep learning theory between *formal, rigorous proofs* (like are traditional in CS) and *approximate, nonrigorous derivations* (e.g. the replica method, scaling arguments, and other physics tricks). Should we care about rigor?
A thread:
Karan Shah retweeted
Our program would not be possible without the support of our awesome reviewers! To honor their hard work, we are excited to announce the Outstanding Reviewer Awards!
Karan Shah retweeted
👇#DigitalTwins of Complex Systems
@CASUSscience Annual Workshop
Dec 6-9 2021
Online using Zoom
casus.science/events/casus-a…
#HPC #AI via @ohm314
On day 1, Veit Hagenmeyer (@KITKarlsruhe) introduces us to #digitaltwins of large-scale #energygrids. (Expect a live demonstration!) Later on, based on #multiphotonimaging, 3D reconstructions & #simulations Marino Zerial from @ZerialLab/@mpicbg tells us how our liver functions.
Karan Shah retweeted
Physicists get a lot of flak for "spherical cows", but computer scientists are on another level.
Karan Shah retweeted
1/ Astronomers made the first clear detection of a moon-forming disc around an exoplanet, using @ALMAObs , in which ESO is a partner.
🔗eso.org/public/news/eso2111/
Credit: ALMA (ESO/NAOJ/NRAO)/Benisty et al.
Karan Shah retweeted
Today with @emblebi, we're launching the #AlphaFold Protein Structure Database, which offers the most complete and accurate picture of the human proteome, doubling humanity’s accumulated knowledge of high-accuracy human protein structures - for free: dpmd.ai/alphafolddb 1/