Great new work by Frank Noe Steven Brunton Andreas Mardt BH Tim Hempel et al Freie Universität Mechanical Engineering at UW UW Engineering CPBF - '#Deeptime: a #Python library for #machinelearning dynamical models from #timeseries...' - bit.ly/3rPRHvn #fluids #compchem #dynamics

Also Yaoyi Chen welcome to Twitter ;-)

New preprint! If at #ACSSpring2022, join me for my talk today: „Multiscale Boltzmann generators for coarse-graining.“ 2:00pm PT. Room 24B. In-person only.

Consider submitting to our research topic Spatiotemporal Investigation of Organization, Aggregation, and Self-Assembly at Membranes frontiersin.org/research-topic… Thanks Mohsen for initiating; co-edited by Thomas Weikl, Shreyas Kaptan, Mauricio J. del Razo, @Blendenfleck, and myself.

Temperature steerable flows and Boltzmann generators, Manuel Dibak, Leon Klein, Andreas Krämer, and Frank Noé Leon Klein Andreas Krämer Frank Noe Freie Universität #StatisticalPhysics go.aps.org/3MARa93

New preprint combining Boltzmann generators and enhanced sampling! arxiv.org/abs/2210.14104 We use normalizing flows to run replica exchange without having to simulate multiple intermediate copies of the systems. We call it _learned replica exchange_ (LREX)

Check out Michele Invernizzi's new preprint on using #BoltzmannGenerators for #MachineLearning enhanced sampling with replica-exchange molecular dynamics.

Free energy practitioners, don't miss this! Co-supervising Maaike's MSc project was a joy. Excited to see her future PhD work.

So happy this is out! 🙂

With Michele Ceriotti, Gabor Csanyi, and Lixin Sun, we are organizing an awesome conference in Berlin on June 19-23: "Bridging length scales with machine learning: from wavefunctions to thermodynamics". Check it out: sites.google.com/view/cecam-psi…

Another manuscript on the ongoing effort of our labs in the development of accurate coarse-grained models for proteins. With Frank Noe, Yaoyi Chen Andreas Krämer Aleks Durumeric, and Nick Charron. Check it out! arxiv.org/abs/2302.07071

I will be opening two fully funded PhD positions (5 year) in my group in the near future. If you are interested in working with generative models in molecular applications (design and simulation) please feel free to reach out.

Soon @ ICML Conference ! 🌪️Normalizing flows for rigid bodies and the rotation group SO(3)! 🌪️ We found a way to design smooth normalizing flows for rigid body systems, like ice 👇 Joint work with amazing Michele Invernizzi Pim de Haan and Frank Noe 1/ arxiv.org/abs/2301.11355

📢Looking for postdocs to join my new group at UMich! We will design molecules with pharmaceutical and environmental relevance through exploration of molecular dynamics Feel free to RT, more details in comments

Leon Klein introduces flow matching with equivariance. This allows us for the first time to get flows working for iid sampling of molecular structures in Cartesian coordinates in such a way that we can get reasonable Monte Carlo acceptance rates. arxiv.org/abs/2306.15030

🚀 Introducing "Equivariant Flow Matching"! Training equivariant Boltzmann generators has been a challenge, but thanks to a tour de force by Leon Klein, we have an efficient method🌟 On a tight timeframe, Leon executed it with modest supervision and landed it in #NeurIPS2023👏

A preprint of the OpenMM 8 paper is out now: x.com/ML_Chem/status…

I am excited to present this work, result of a 4-year big collaborative project: arxiv.org/abs/2310.18278 #MachineLearning a transferable bottom-up protein force field, trained on force data from over all-atom MD simulations, using physical priors and graph neural networks.🧵⬇️

Hello again 🇺🇸 🏙️ Today was my first day as Machine Learning Researcher at D. E. Shaw Research. I am very excited to be joining an awesome team developing computational tools to transform drug discovery! And it’s unreal that I’ll get to enjoy this view every day now🗽

I am at #NeurIPS23 all week, presenting our work Equivariant Flow Matching and Timewarp (more details below). Swing by if you're curious, and feel free to send a message if you'd like to connect or meet and discuss generative models for molecular sampling.