Jonas Köhler
@jonkhler.argmin.xyz
web @ https://argmin.xyz
interests: machine learning, ai4science, algorithms, coding
member of technical staff @ https://cusp.ai
past @ MSR, DeepMind, MPI-IS
home @ Heimbach (Gilserberg), Berlin, Europe
born @ 353 ppm
block toxicity
he/him
interests: machine learning, ai4science, algorithms, coding
member of technical staff @ https://cusp.ai
past @ MSR, DeepMind, MPI-IS
home @ Heimbach (Gilserberg), Berlin, Europe
born @ 353 ppm
block toxicity
he/him
Reposted by Jonas Köhler
We have a job opening at Peptone.io for an ML researcher.
Come help us find new ways to understand and drug intrinsically disordered proteins (IDPs), it's a very interesting and important problem!
Link in the reply ↓
Come help us find new ways to understand and drug intrinsically disordered proteins (IDPs), it's a very interesting and important problem!
Link in the reply ↓
May 9, 2025 at 3:26 PM
We have a job opening at Peptone.io for an ML researcher.
Come help us find new ways to understand and drug intrinsically disordered proteins (IDPs), it's a very interesting and important problem!
Link in the reply ↓
Come help us find new ways to understand and drug intrinsically disordered proteins (IDPs), it's a very interesting and important problem!
Link in the reply ↓
Spending the Easter days prepping for #ICLR2025 in #singapore 😍🤩
If you want to meet/catch up on
#CuspAI
#AI4Science #MachineLearning and/or ML for #materials / #simulation ping me.!
DMs are open :)
My lovely colleagues Sina, @sindylowe.bsky.social and Govert are around too!
If you want to meet/catch up on
#CuspAI
#AI4Science #MachineLearning and/or ML for #materials / #simulation ping me.!
DMs are open :)
My lovely colleagues Sina, @sindylowe.bsky.social and Govert are around too!
April 19, 2025 at 8:04 PM
Spending the Easter days prepping for #ICLR2025 in #singapore 😍🤩
If you want to meet/catch up on
#CuspAI
#AI4Science #MachineLearning and/or ML for #materials / #simulation ping me.!
DMs are open :)
My lovely colleagues Sina, @sindylowe.bsky.social and Govert are around too!
If you want to meet/catch up on
#CuspAI
#AI4Science #MachineLearning and/or ML for #materials / #simulation ping me.!
DMs are open :)
My lovely colleagues Sina, @sindylowe.bsky.social and Govert are around too!
Reposted by Jonas Köhler
The blog post of the late Felix Hill is powerful. Stress for AI researchers today is real.
I did not know Felix Hill and I am sorry for those who did.
This story is perhaps a reminder for students, postdocs, founders and researchers to take care of their well being.
medium.com/@felixhill/2...
I did not know Felix Hill and I am sorry for those who did.
This story is perhaps a reminder for students, postdocs, founders and researchers to take care of their well being.
medium.com/@felixhill/2...
200bn Weights of Responsibility
The Stress of Working in Modern AI
medium.com
January 4, 2025 at 10:44 AM
The blog post of the late Felix Hill is powerful. Stress for AI researchers today is real.
I did not know Felix Hill and I am sorry for those who did.
This story is perhaps a reminder for students, postdocs, founders and researchers to take care of their well being.
medium.com/@felixhill/2...
I did not know Felix Hill and I am sorry for those who did.
This story is perhaps a reminder for students, postdocs, founders and researchers to take care of their well being.
medium.com/@felixhill/2...
Unreal landscapes ❤️
December 31, 2024 at 1:33 PM
Unreal landscapes ❤️
@imtd.bsky.social convinced me to turn my homepage into a blog.
Not much content yet... Let's see if I stick to it: argmin.xyz.
Not much content yet... Let's see if I stick to it: argmin.xyz.
December 26, 2024 at 12:06 PM
@imtd.bsky.social convinced me to turn my homepage into a blog.
Not much content yet... Let's see if I stick to it: argmin.xyz.
Not much content yet... Let's see if I stick to it: argmin.xyz.
And today, happy Hanukkah 🕎❤️
December 25, 2024 at 1:19 PM
And today, happy Hanukkah 🕎❤️
Fun challenge that is apparently hard for LLMs (at least all those I could get my hands on): solving carefully drafted anagrams!
New ARC challenge? :D
PS: credits go to family Lehmann from Kassel who invent those ridiculously hard Xmas puzzles every year!
New ARC challenge? :D
PS: credits go to family Lehmann from Kassel who invent those ridiculously hard Xmas puzzles every year!
December 25, 2024 at 11:10 AM
Fun challenge that is apparently hard for LLMs (at least all those I could get my hands on): solving carefully drafted anagrams!
New ARC challenge? :D
PS: credits go to family Lehmann from Kassel who invent those ridiculously hard Xmas puzzles every year!
New ARC challenge? :D
PS: credits go to family Lehmann from Kassel who invent those ridiculously hard Xmas puzzles every year!
Merry Christmas! ❤️🌲
December 24, 2024 at 1:54 PM
Merry Christmas! ❤️🌲
Segment radixsort, prefixsums and pointwise transforms are just OPed on GPUs.
The amount of throughput you can get on a single cheap device is ridiculous! And you can implement so much crazy stuff with just that. ❤️
The amount of throughput you can get on a single cheap device is ridiculous! And you can implement so much crazy stuff with just that. ❤️
December 20, 2024 at 5:25 PM
Segment radixsort, prefixsums and pointwise transforms are just OPed on GPUs.
The amount of throughput you can get on a single cheap device is ridiculous! And you can implement so much crazy stuff with just that. ❤️
The amount of throughput you can get on a single cheap device is ridiculous! And you can implement so much crazy stuff with just that. ❤️
Reposted by Jonas Köhler
hello bluesky! we have a new preprint on solvation free energies:
tl;dr: We define an interpolating density by its sampling process, and learn the corresponding equilibrium potential with score matching. arxiv.org/abs/2410.15815
with @francois.fleuret.org and @tbereau.bsky.social
(1/n)
tl;dr: We define an interpolating density by its sampling process, and learn the corresponding equilibrium potential with score matching. arxiv.org/abs/2410.15815
with @francois.fleuret.org and @tbereau.bsky.social
(1/n)
December 17, 2024 at 12:32 PM
hello bluesky! we have a new preprint on solvation free energies:
tl;dr: We define an interpolating density by its sampling process, and learn the corresponding equilibrium potential with score matching. arxiv.org/abs/2410.15815
with @francois.fleuret.org and @tbereau.bsky.social
(1/n)
tl;dr: We define an interpolating density by its sampling process, and learn the corresponding equilibrium potential with score matching. arxiv.org/abs/2410.15815
with @francois.fleuret.org and @tbereau.bsky.social
(1/n)
Awesome work! Amazing to see this out :)
Super excited to preprint our work on developing a Biomolecular Emulator (BioEmu): Scalable emulation of protein equilibrium ensembles with generative deep learning from @msftresearch.bsky.social ch AI for Science.
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
December 6, 2024 at 9:48 AM
Awesome work! Amazing to see this out :)
Reposted by Jonas Köhler
Reposted by Jonas Köhler
I’ll be at NeurIPS next week –happy to connect! Hit me up if you’d like to meet up and chat :)
(tabular deep learning; in-context learning; SSL; AI in EU; politics; where to get good coffee/food; your favourite Parks and Rec moment; …)
(tabular deep learning; in-context learning; SSL; AI in EU; politics; where to get good coffee/food; your favourite Parks and Rec moment; …)
December 2, 2024 at 7:45 AM
I’ll be at NeurIPS next week –happy to connect! Hit me up if you’d like to meet up and chat :)
(tabular deep learning; in-context learning; SSL; AI in EU; politics; where to get good coffee/food; your favourite Parks and Rec moment; …)
(tabular deep learning; in-context learning; SSL; AI in EU; politics; where to get good coffee/food; your favourite Parks and Rec moment; …)
Every two days I get this "invalid handle" label for my custom handle @jonkhler.argmin.xyz (using the DNS TXT config).
This seems to be a bug - whenever I reset it, it's fine for another bunch of days.
Anyone else experiencing this? If so, what's the cause/fix?
#bluesky @jay.bsky.team
This seems to be a bug - whenever I reset it, it's fine for another bunch of days.
Anyone else experiencing this? If so, what's the cause/fix?
#bluesky @jay.bsky.team
December 1, 2024 at 11:05 AM
Every two days I get this "invalid handle" label for my custom handle @jonkhler.argmin.xyz (using the DNS TXT config).
This seems to be a bug - whenever I reset it, it's fine for another bunch of days.
Anyone else experiencing this? If so, what's the cause/fix?
#bluesky @jay.bsky.team
This seems to be a bug - whenever I reset it, it's fine for another bunch of days.
Anyone else experiencing this? If so, what's the cause/fix?
#bluesky @jay.bsky.team
Reposted by Jonas Köhler
I wrote a summary of the main ingredients of the neat proof by Hugo Lavenant that diffusion models do not generally define optimal transport. github.com/mathematical...
November 30, 2024 at 8:35 AM
I wrote a summary of the main ingredients of the neat proof by Hugo Lavenant that diffusion models do not generally define optimal transport. github.com/mathematical...
Reposted by Jonas Köhler
The link between diffusion models and optimal transport is still a bit of an enigma to me.
One thing that's clear: different diffusion models trained on similar datasets tend to recover similar mappings. If these are generally not OT, in what sense are they optimal instead?
One thing that's clear: different diffusion models trained on similar datasets tend to recover similar mappings. If these are generally not OT, in what sense are they optimal instead?
I wrote a summary of the main ingredients of the neat proof by Hugo Lavenant that diffusion models do not generally define optimal transport. github.com/mathematical...
November 30, 2024 at 12:56 PM
The link between diffusion models and optimal transport is still a bit of an enigma to me.
One thing that's clear: different diffusion models trained on similar datasets tend to recover similar mappings. If these are generally not OT, in what sense are they optimal instead?
One thing that's clear: different diffusion models trained on similar datasets tend to recover similar mappings. If these are generally not OT, in what sense are they optimal instead?
Reposted by Jonas Köhler
I find strange that engineering the gradient radically to learn better is not a whole field.
November 29, 2024 at 3:59 AM
I find strange that engineering the gradient radically to learn better is not a whole field.
Reposted by Jonas Köhler
November 28, 2024 at 11:12 PM
cuda level programming is fun - but also makes me curse a lot :D
November 28, 2024 at 6:28 PM
cuda level programming is fun - but also makes me curse a lot :D
Reposted by Jonas Köhler
I mean, we all know who the only true gold standard for AI publications is, and by that measure it’s not looking good!
November 28, 2024 at 10:31 AM
I mean, we all know who the only true gold standard for AI publications is, and by that measure it’s not looking good!
Reposted by Jonas Köhler
Reposted by Jonas Köhler
Super excited and proud of the new additions to Aurora! Check them out here :)
🎉Excited to announce major new breakthroughs in our Aurora foundation model! Our team (@cbodnar.com, @wessel.ai, @megstanley.bsky.social, @a-lucic.bsky.social, Anna Vaughan) has achieved unprecedented results across multiple Earth system forecasting tasks. Here's what's new... 🧵
November 27, 2024 at 9:37 PM
Super excited and proud of the new additions to Aurora! Check them out here :)
Hm. Seems like deep learning is approaching the wall 🤔
November 28, 2024 at 9:24 AM
Hm. Seems like deep learning is approaching the wall 🤔