Li Li

My last work at @GoogleAI "Evolving symbolic density functionals" just appeared @ScienceAdvances !Inspired by recent advances in AutoML and program synthesis, we show that ML can compose symbolic density functionals with simplicity as human-designed ones. science.org/doi/10.1126/sc…

@jhrmnn @ja_kirkpatrick @FU_Berlin @bifoldberlin Congratulations!

Solving Schrödinger Equation Using Tensor Neural Network arxiv.org/abs/2209.12572 #compchem

Very interesting! A bit like Autopilot but for your computer.

Learning with Differentiable Algorithms If you’re interested in making algorithms differentiable so that algorithms can be learned from training data, @FHKPetersen’s thesis covers differentiable versions of Sorting and Ranking, Top-k, Rendering, Logic. arxiv.org/abs/2209.00616

@MilesCranmer This is a nice one

Just discovered this 9-digit approximation to π with PySR's genetic algorithm... is this a known formula?

My friend @jure claims that modern #recsys at most companies are "based on graphical representations" (youtu.be/aBHC6xzx9YI min.8). In my experience that is not the case. Graphs might be *used*, but are never the fundamental construct. Anyone has similar/different experience?

@tunguz hello :-)

@leeley18 oh hi

Most people will be in the exact same place next year as they are today. Don’t be most people.

Fei-Fei to me after I showed her my first image captioning (image to text) network around 2015: “very cool, now do it backwards!”. Me: “haha that’s impossible” 🥲. Turns out you just need a few ~B alt-text dataset scrape, transformer, diffusion, and a cluster of ~thousand A100s.

Yesterday @nnaisense released EvoTorch (evotorch.ai), a state-of-the-art evolutionary algorithm library built on @PyTorch, with GPU-acceleration and easy training on huge compute clusters using @raydistributed. (1/2)

@shoyer @DynamicWebPaige Yes! I am glad that it is official. This can unlock a lot of interesting applications!

@DynamicWebPaige @leeley18 in case you haven't seen this one!

✨🧪 "PySCF with auto-diff provides a simple, lightweight, and efficient platform for quantum chemistry calculations and methodology—and can be used to simulate the properties of molecules, crystals, and custom Hamiltonians." 📄arxiv.org/abs/2207.13836 👩‍💻github.com/fishjojo/pyscf…

@tunguz A handcrafted neural network in C++ in @DanielWhiteson numerical computation class project.

What was the first ML algorithm, aside from the Logistic/Linear Regression, that you ever used? Mine was a simple hand-crafted decision tree.

@quantumVerd sure thing

By age 30 you should have fully trained a QNN on a real quantum device

@olexandr @rbhar90 Do you mean that geometric deep-learning does not have any advantage over direct deep learning?

@rbhar90 I did *NOT* say that;) New models are so flexible that could learn almost anything.

geometric deep-learning that does fast molecular docking is the biggest scam since Theranos

A nice article from Dr. Luciano Abriata featuring our recent work on symbolic regression and density functional theory done at @GoogleAI towardsdatascience.com/google-propose… #dft #physics #ai4science #ml Link to our paper "Evolving symbolic density functionals" arxiv.org/abs/2203.02540

@TheNerdStation what is GCP then?

There's a special place in hell for people who code like this: if condition == True: return True else: return False Don't be one of those people!

levels.fyi is your friend

@AbhijeetJK Thanks for your interest.

@leeley18 Super interesting🧠

Can computer construct density functionals in the symbolic form? I would like to share with you our latest work on machine learning applying to density functional theory -- Evolving symbolic density functionals (arxiv.org/abs/2203.02540) #SymbolicRegression #ProgramSynthesis 1/3