Pinned
Jonas Köhler
2,812 posts
Jonas Köhler
@jonkhler
fueled by randomness
ml research @cusp_ai, ex-{@MSFTResearch,@GoogleDeepMind,@MPI_IS}
born @ 353 ppm, he/him, block toxicity
- Now it is on me to send off such a Tweet: Hugely excited to announce that I will be joining @DeepMind's Science team for a Research Scientist internship starting in September! I am humbled - and thrilled about the topics I am going to work on! :)
- Cool list on equivariant networks! If you think some work is missing, please make a PR! :-)
- Soon @ @icmlconf ! 🌪️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 @inve_michele @pimdehaan and @FrankNoeBerlin 1/ arxiv.org/abs/2301.11355
GIF - Using this occasion to lift the curtain as well :) I will stay in Berlin and join @MSFTResearch/@FrankNoeBerlin end of the year as a Senior Researcher, Deep Learning for Molecular Modeling 🧬🤖📎 Very excited about this opportunity and very happy to continue the joint work ❤️Tweeps - news! I'll join @MSFTResearch as research manager starting Oct 1st, opening a new AI4Science Lab in beautiful Berlin! We'll focus on fundamental challenges between #MachineLearning & #Physics/#Chemistry. We'll start at Alexanderplatz, and we'll hire at all levels! 🧵
- Excited to join @cusp_ai as a member of technical staff to help accelerate material discovery with machine learning research. Our goal is to find improved materials for carbon capture - a mission very close to my heart. Let's clean this up! 💪❤️ #MachineLearning
- NeurIPS reviewers are like: "Why does your model for molecular systems not work on images?! What is the point of such model??" That's not fun. :(
- Replying to @aravindr93Thanks God, there are a lot of good places outside the US where good students with curiosity and ambition can still do excellent ML research and where they do not "need a PhD to get accepted". Stories like these ignite anxiety among graduates and are not very helpful imho.
- reddit/ml has an increasingly worse signal-to-noise ratio (+ a lot of toxicity problems...) and Twitter is not really well suited for discussions. So do people have recommendations for discussing ML research online beyond the formalized context of a conference? #MachineLearning
- A bit late to the party... ...but we are happy to announce that our work on Equivariant Normalizing Flows which are able to sample from symmetric target distributions has been accepted to ICML 2020! arxiv.org/abs/2006.02425 Joint work with #LeonKlein @FrankNoeBerlin 1/
- After last week's tweet: here the promised preprint. Happy to announce, that our paper on Smooth Normalizing Flows (w/ @__kraemer__ @FrankNoeBerlin) got accepted to #NeurIPS 2021! Link: arxiv.org/abs/2110.00351 1/
- Absolute gem of a paper to understand what I actually did in the past 😅 arxiv.org/abs/2401.14131
- Replying to @jonkhler and @aravindr93PS: I read your tweet in good spirit, but breaking the myth that someone needs to graduate with a PhD from Stanford in order to meaningfully contribute to ML research is critical: it is (a) most likely not true and (b) might lead less confident student to not pursue their dream.
- 2006: semi-prof reggae band 2008: politics & economics 2009: pure math 2011: major depression 2012: art school 2013: 20h/w webdev 2016: bsc applied cs 2017: married 2018: msc AI 2018: start PhD in ML+Physics 2020: beat depression 2021: DeepMind internship 2023: first real job2013: married 2014: Cirque du Soleil employee 2015: alpaca rancher 2016: masters in finance 2017: first and last time with a Real Job 2018: started PhD 2019: PhD going good 2020: PhD going bad 2021: divorced 2022: get a doctorate 2023: stanford person 2024: startup goober
