Some non-MOF content today to clean your palate.
Want 70x fast, 2x accurate Hessians for any equivariant machine learning force field? We got you!
#compchemsky #chemsky #machinelearning #mlff
Want 70x fast, 2x accurate Hessians for any equivariant machine learning force field? We got you!
#compchemsky #chemsky #machinelearning #mlff
What do transition state search, geometry relaxation, zero point energy corrections, and extrema classification have in common? They all require Hessians!
The problem is, accurate Hessians are really expensive, even with MLIPs. We say, just shoot 'em from the HIP! 🤠
[1/4]
The problem is, accurate Hessians are really expensive, even with MLIPs. We say, just shoot 'em from the HIP! 🤠
[1/4]
October 8, 2025 at 6:14 PM
Some non-MOF content today to clean your palate.
Want 70x fast, 2x accurate Hessians for any equivariant machine learning force field? We got you!
#compchemsky #chemsky #machinelearning #mlff
Want 70x fast, 2x accurate Hessians for any equivariant machine learning force field? We got you!
#compchemsky #chemsky #machinelearning #mlff
🚀 We asked: What if the time to set up and run the best Machine-Learned Interatomic Potentials (MLIPs) took seconds, not days?
Today, we release the MLIP Garden v0.1.
What you can do now:
- Experiment ~instantly
- Scale deployments on experimental NSF and Dept of Energy systems
Today, we release the MLIP Garden v0.1.
What you can do now:
- Experiment ~instantly
- Scale deployments on experimental NSF and Dept of Energy systems
August 20, 2025 at 7:53 PM
Our First Ever Video in 2026 !!
Full of fun moments and memes as always
Live on my YT channel Please enjoy 🧡
www.youtube.com/watch?v=MLfF...
Full of fun moments and memes as always
Live on my YT channel Please enjoy 🧡
www.youtube.com/watch?v=MLfF...
My Cat Is Kinda Homeless !! | VRChat Moments
YouTube video by Catzy
www.youtube.com
January 4, 2026 at 7:10 PM
Our First Ever Video in 2026 !!
Full of fun moments and memes as always
Live on my YT channel Please enjoy 🧡
www.youtube.com/watch?v=MLfF...
Full of fun moments and memes as always
Live on my YT channel Please enjoy 🧡
www.youtube.com/watch?v=MLfF...
Ever get tired of tiny timesteps bottlenecking your MD simulations?
We show how to train a model for large-timestep Hamiltonian dynamics directly on standard MLFF datasets. 𝗡𝗼 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝘁𝗿𝗮𝗷𝗲𝗰𝘁𝗼𝗿𝗶𝗲𝘀, 𝗻𝗼 𝘂𝗻𝗿𝗼𝗹𝗹𝗶𝗻𝗴, 𝗻𝗼 𝘁𝗲𝗮𝗰𝗵𝗲𝗿 needed!
🧵👇
We show how to train a model for large-timestep Hamiltonian dynamics directly on standard MLFF datasets. 𝗡𝗼 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝘁𝗿𝗮𝗷𝗲𝗰𝘁𝗼𝗿𝗶𝗲𝘀, 𝗻𝗼 𝘂𝗻𝗿𝗼𝗹𝗹𝗶𝗻𝗴, 𝗻𝗼 𝘁𝗲𝗮𝗰𝗵𝗲𝗿 needed!
🧵👇
February 19, 2026 at 3:05 PM
Ever get tired of tiny timesteps bottlenecking your MD simulations?
We show how to train a model for large-timestep Hamiltonian dynamics directly on standard MLFF datasets. 𝗡𝗼 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝘁𝗿𝗮𝗷𝗲𝗰𝘁𝗼𝗿𝗶𝗲𝘀, 𝗻𝗼 𝘂𝗻𝗿𝗼𝗹𝗹𝗶𝗻𝗴, 𝗻𝗼 𝘁𝗲𝗮𝗰𝗵𝗲𝗿 needed!
🧵👇
We show how to train a model for large-timestep Hamiltonian dynamics directly on standard MLFF datasets. 𝗡𝗼 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝘁𝗿𝗮𝗷𝗲𝗰𝘁𝗼𝗿𝗶𝗲𝘀, 𝗻𝗼 𝘂𝗻𝗿𝗼𝗹𝗹𝗶𝗻𝗴, 𝗻𝗼 𝘁𝗲𝗮𝗰𝗵𝗲𝗿 needed!
🧵👇
www.youtube.com/watch?v=MLFF...
Something I have had planned out and am happy to present, a video further teasing the story of my Puyo Puyo AU!
Something I have had planned out and am happy to present, a video further teasing the story of my Puyo Puyo AU!
My Puyo Puyo AU Introduction
YouTube video by warelander
www.youtube.com
January 10, 2025 at 3:50 PM
www.youtube.com/watch?v=MLFF...
Something I have had planned out and am happy to present, a video further teasing the story of my Puyo Puyo AU!
Something I have had planned out and am happy to present, a video further teasing the story of my Puyo Puyo AU!
Does anyone have the story behind the MLFF (machine learning force field) vs MLIP (machine learning interatomic potential) naming split? #compchem
November 13, 2024 at 8:30 PM
Does anyone have the story behind the MLFF (machine learning force field) vs MLIP (machine learning interatomic potential) naming split? #compchem
#REMTop30 - Bonus entry
#MusicChallenge
Yesterday we learned that the traitor Oliver North had married his secretary from the Iran-Contra era Fawn Hall. North is namechecked in REM's 1995 track Revolution...
www.youtube.com/watch?v=MlFF...
#MusicChallenge
Yesterday we learned that the traitor Oliver North had married his secretary from the Iran-Contra era Fawn Hall. North is namechecked in REM's 1995 track Revolution...
www.youtube.com/watch?v=MlFF...
R.E.M. - Revolution (Road Movie '95)
YouTube video by REMfanfrance
www.youtube.com
September 10, 2025 at 9:46 AM
#REMTop30 - Bonus entry
#MusicChallenge
Yesterday we learned that the traitor Oliver North had married his secretary from the Iran-Contra era Fawn Hall. North is namechecked in REM's 1995 track Revolution...
www.youtube.com/watch?v=MlFF...
#MusicChallenge
Yesterday we learned that the traitor Oliver North had married his secretary from the Iran-Contra era Fawn Hall. North is namechecked in REM's 1995 track Revolution...
www.youtube.com/watch?v=MlFF...
Our new preprint is out: Extending the RANGE of Graph Neural Networks: Relaying Attention Nodes for Global Encoding.
arxiv.org/abs/2502.13797
We present a new architecture to address the problem of "short-sightedness" of MLFF based on GNN, which outperform in speed and accuracy Ewald-based GNN.
arxiv.org/abs/2502.13797
We present a new architecture to address the problem of "short-sightedness" of MLFF based on GNN, which outperform in speed and accuracy Ewald-based GNN.
Extending the RANGE of Graph Neural Networks: Relaying Attention Nodes for Global Encoding
Graph Neural Networks (GNNs) are routinely used in molecular physics, social sciences, and economics to model many-body interactions in graph-like systems. However, GNNs are inherently local and can s...
arxiv.org
February 20, 2025 at 10:10 AM
Our new preprint is out: Extending the RANGE of Graph Neural Networks: Relaying Attention Nodes for Global Encoding.
arxiv.org/abs/2502.13797
We present a new architecture to address the problem of "short-sightedness" of MLFF based on GNN, which outperform in speed and accuracy Ewald-based GNN.
arxiv.org/abs/2502.13797
We present a new architecture to address the problem of "short-sightedness" of MLFF based on GNN, which outperform in speed and accuracy Ewald-based GNN.
For a MLFF, I’d think Cr2 isn’t really much harder than H2. If it’s in training data, it’s easy. If not, it seems impossible to guess from any other bond dissociation. Smoothness is a whole another issue of course
November 18, 2024 at 7:41 AM
For a MLFF, I’d think Cr2 isn’t really much harder than H2. If it’s in training data, it’s easy. If not, it seems impossible to guess from any other bond dissociation. Smoothness is a whole another issue of course
Force fields with QM accuracy and MM speed are the theoretical biophysicist's Philosopher's Stone. I used ANIX2 with OpenFF and OpenMM. The results were insightful but at the expense of huge computational costs. #MLFF #biophysics #moleculardynamics
On the design space between molecular mechanics and machine learning force fields
A force field as accurate as quantum mechanics (QMs) and as fast as molecular mechanics (MMs), with which one can simulate a biomolecular system efficiently eno
pubs.aip.org
April 5, 2025 at 10:48 AM
Force fields with QM accuracy and MM speed are the theoretical biophysicist's Philosopher's Stone. I used ANIX2 with OpenFF and OpenMM. The results were insightful but at the expense of huge computational costs. #MLFF #biophysics #moleculardynamics
I tend to use "interatomic potential" when referring to a particular functional form, such as Lennard-Jones, Buckingham, Morse... "Force field" is more flexible, so I sit in camp #MLFF
November 14, 2024 at 8:22 PM
I tend to use "interatomic potential" when referring to a particular functional form, such as Lennard-Jones, Buckingham, Morse... "Force field" is more flexible, so I sit in camp #MLFF
An excellent read for MLFF/MLIP people: "Benchmarking of Universal Machine Learning Interatomic Potentials for Structural Relaxation"
openreview.net/pdf?id=fNyXC...
Simple test sets of structures are not enough. A useful model needs to be able to start from a bad structure and find a good one!
openreview.net/pdf?id=fNyXC...
Simple test sets of structures are not enough. A useful model needs to be able to start from a bad structure and find a good one!
November 14, 2024 at 4:51 PM
An excellent read for MLFF/MLIP people: "Benchmarking of Universal Machine Learning Interatomic Potentials for Structural Relaxation"
openreview.net/pdf?id=fNyXC...
Simple test sets of structures are not enough. A useful model needs to be able to start from a bad structure and find a good one!
openreview.net/pdf?id=fNyXC...
Simple test sets of structures are not enough. A useful model needs to be able to start from a bad structure and find a good one!
“Kapsch will either upgrade existing tolling infrastructure or build new roadside systems along previously untolled roads”
www.itsinternational.com/news/kapsch-...
www.itsinternational.com/news/kapsch-...
Kapsch signs €7.5m Norway free-flow tolling deal
Vegfinans will deploy Autopass MLFF across five counties
www.itsinternational.com
February 21, 2025 at 5:39 AM
“Kapsch will either upgrade existing tolling infrastructure or build new roadside systems along previously untolled roads”
www.itsinternational.com/news/kapsch-...
www.itsinternational.com/news/kapsch-...
Just another data point: the upcoming ORCA 6.1 (release next Tuesday 17/06!) will come with a simpler interface for all kinds of MLFF, including a tutorial for the UMA.
One can use it to optimize geometries, run GOAT, NEB or anything else.
Hope it helps!
One can use it to optimize geometries, run GOAT, NEB or anything else.
Hope it helps!
June 12, 2025 at 11:28 AM
Just another data point: the upcoming ORCA 6.1 (release next Tuesday 17/06!) will come with a simpler interface for all kinds of MLFF, including a tutorial for the UMA.
One can use it to optimize geometries, run GOAT, NEB or anything else.
Hope it helps!
One can use it to optimize geometries, run GOAT, NEB or anything else.
Hope it helps!
Touch ‘n Go, STAR Systems International Demonstrate MLFF Toll System #mlff #tng #touchngo #transportation
Touch ‘n Go, STAR Systems International Demonstrate MLFF Toll System
The Multi-Lane Fast Flow (MLFF) system remains a subject of interest in the country. Recently, Touch ’n Go Sdn Bhd (TnG) announced its collaboration with STAR Systems International (SSI) to demonst…
www.lowyat.net
October 8, 2025 at 6:00 AM
Touch ‘n Go, STAR Systems International Demonstrate MLFF Toll System #mlff #tng #touchngo #transportation
A persistent challenge in atomistic modelling is combining quantum-level accuracy with the efficiency needed for large, complex simulations. Our pretrained lightweight SO3LR MLFF is a step toward this goal.
September 7, 2025 at 4:24 PM
A persistent challenge in atomistic modelling is combining quantum-level accuracy with the efficiency needed for large, complex simulations. Our pretrained lightweight SO3LR MLFF is a step toward this goal.
Waiting for your MD simulation to finish? Visit our spotlight paper at ICML on how to train large timestep models on existing MLFF datasets in the meantime!
🗓️ Wednesday, July 8
⏰ 10:30 AM – 12:15 PM KST
arxiv.org/abs/2601.22123
🗓️ Wednesday, July 8
⏰ 10:30 AM – 12:15 PM KST
arxiv.org/abs/2601.22123
Excited to be at ICML with @plainer.bsky.social & @gregorlied.bsky.social to present our 🌟 spotlight 🌟 paper on Hamiltonian Flow Maps for large-timestep molecular dynamics!
Come say hi at our poster tomorrow or send us a message if you want to chat! 🤗
🗓️ July 8, 10:30 AM – 12:15 PM
📍 HALL A #1306
Come say hi at our poster tomorrow or send us a message if you want to chat! 🤗
🗓️ July 8, 10:30 AM – 12:15 PM
📍 HALL A #1306
July 7, 2026 at 12:48 PM
Waiting for your MD simulation to finish? Visit our spotlight paper at ICML on how to train large timestep models on existing MLFF datasets in the meantime!
🗓️ Wednesday, July 8
⏰ 10:30 AM – 12:15 PM KST
arxiv.org/abs/2601.22123
🗓️ Wednesday, July 8
⏰ 10:30 AM – 12:15 PM KST
arxiv.org/abs/2601.22123
My top 3 new papers in @jacs.acspublications.org:
i) Recipe for Flat Bands in Pyrochlore Materials
pubs.acs.org/doi/10.1021/...
ii) MACE-OFF... MLFF for Organic Molecules pubs.acs.org/doi/10.1021/...
iii) Ptychographic Observation of Li... at Sub-Angstrom Resolution pubs.acs.org/doi/10.1021/...
i) Recipe for Flat Bands in Pyrochlore Materials
pubs.acs.org/doi/10.1021/...
ii) MACE-OFF... MLFF for Organic Molecules pubs.acs.org/doi/10.1021/...
iii) Ptychographic Observation of Li... at Sub-Angstrom Resolution pubs.acs.org/doi/10.1021/...
Recipe for Flat Bands in Pyrochlore Materials: A Chemist’s Perspective
Materials in which atoms are arranged in a pyrochlore lattice have found renewed interest, as, at least theoretically, orbitals on that lattice can form flat bands. However, real materials often do not behave according to theoretical models, which is why there has been a dearth of pyrochlore materials exhibiting flat band physics. Here, we examine the conditions under which ideal “pyrochlore bands” can exist in real materials and how to have those close to the Fermi level. We find that the simple model used in the literature does not apply to the bands at the Fermi level in real pyrochlore materials. However, surprisingly, we find that certain oxide compounds that have oxygen orbitals inside the pyrochlore tetrahedra do exhibit near-ideal pyrochlore bands near the Fermi level. We explain this observation by a generalized tight-binding model, including the oxygen orbitals. We further classify all known pyrochlore materials based on their crystal structure, band structure, and chemical characteristics and propose materials to study in future experiments.
pubs.acs.org
May 20, 2025 at 10:24 AM
My top 3 new papers in @jacs.acspublications.org:
i) Recipe for Flat Bands in Pyrochlore Materials
pubs.acs.org/doi/10.1021/...
ii) MACE-OFF... MLFF for Organic Molecules pubs.acs.org/doi/10.1021/...
iii) Ptychographic Observation of Li... at Sub-Angstrom Resolution pubs.acs.org/doi/10.1021/...
i) Recipe for Flat Bands in Pyrochlore Materials
pubs.acs.org/doi/10.1021/...
ii) MACE-OFF... MLFF for Organic Molecules pubs.acs.org/doi/10.1021/...
iii) Ptychographic Observation of Li... at Sub-Angstrom Resolution pubs.acs.org/doi/10.1021/...