iiinova — where your digital self keeps evolving

Felix Koehler @felix-m-koehler.bsky.social · 04/08/2026

In June 2026 I gave a lecture and three practical sessions on the hybridization of neural networks and numerical solvers. The recordings are now online: www.youtube.com/watch?v=carw...

youtube.com

Hybridization of Neural Networks and Numerical Solvers in JAX with Differentiable Physics

YouTube video by Machine Learning & Simulation

120

Felix Koehler @felix-m-koehler.bsky.social · 14/11/2025

Can your AI surpass the simulator that taught it? What if the key to more accurate PDE modeling lies in questioning your training data's origins? 🤔 Excited to share my #NeurIPS 2025 paper with @thuereygroup.bsky.social: "Neural Emulator Superiority"!

193

Felix Koehler @felix-m-koehler.bsky.social · 19/09/2025

📢 Calling for (4-page) workshop papers on hashtag #differentiable programming and hashtag #SciML @euripsconf.bsky.social.

010

Reposted by Felix Koehler

NeurIPS Europe @neuripseurope.bsky.social · 19/09/2025

It may only be a band-aid, but we have just announced our new "Salon des Refusés" sessions for papers rejected due to space constraints: bsky.app/profile/euri...

4145

Felix Koehler @felix-m-koehler.bsky.social · 09/06/2025

Check Out my latest video on implementing an attention-based neural operator/emulator (i.e. a Transformer) in JAX: youtu.be/GVVWpyvXq_s

youtu.be

Transformer Neural Operator in JAX

YouTube video by Machine Learning & Simulation

130

Felix Koehler @felix-m-koehler.bsky.social · 18/04/2025

Travelling to Singapore next week for #ICLR2025 presenting this paper (Sat 3 pm nr. 538): arxiv.org/abs/2502.19611 DM me (Whova, Email or bsky) if you want to chat about (autoregressive) neural emulators/operators for PDE, autodiff, differentiable physics, numerical solvers etc. 😊

arxiv.org

PRDP: Progressively Refined Differentiable Physics

The physics solvers employed for neural network training are primarily iterative, and hence, differentiating through them introduces a severe computational burden as iterations grow large. Inspired by...

021

Felix Koehler @felix-m-koehler.bsky.social · 04/04/2025

Check out my latest video on approximating the full Lyapunov spectrum for the Lorenz system: youtu.be/Enves8MDwms Nice showcase of #JAX's features: - `jax.lax.scan` for autoregressive rollout - `jax.linearize` repeated jvp - `jax.vmap`: automatic vectorization

youtu.be

Full Lyapunov Spectrum of Chaotic Lorenz System using JAX

YouTube video by Machine Learning & Simulation

100

Felix Koehler @felix-m-koehler.bsky.social · 18/02/2025

Today, I had the chance to present my #NeurIPS paper "APEBench" @SimAI4Science . You can find the recording on YouTube: youtu.be/wie-SzD6AJE

youtu.be

APEBench Talk @ Pasteur Labs Journal Club

YouTube video by Machine Learning & Simulation

051

Felix Koehler @felix-m-koehler.bsky.social · 12/02/2025

Thanks @munichcenterml.bsky.social for highlighting my recent #NeurIPS paper: APEBench, a new benchmark suite for autoregressive emulators of PDEs to understand how we might solve the models of nature more efficiently. More details 🧵 Visual summary on project page: tum-pbs.github.io/apebench-pap...

tum-pbs.github.io

APEBench: A Benchmark for Autoregressive Neural Emulators of PDEs

100

Reposted by Felix Koehler

Roy Frostig @froystig.bsky.social · 04/02/2025

Our online book on systems principles of LLM scaling is live at jax-ml.github.io/scaling-book/ We hope that it helps you make the most of your computing resources. Enjoy!

jax-ml.github.io

How To Scale Your Model

Training LLMs often feels like alchemy, but understanding and optimizing the performance of your models doesn't have to. This book aims to demystify the science of scaling language models on TPUs: how...

3359

Reposted by Felix Koehler

N. Thuerey's research group at TUM @thuereygroup.bsky.social · 23/01/2025

I’d like to thank everyone contributing to our five accepted ICLR papers for the hard work! Great job everyone 👍 Here’s a quick list, stay tuned for details & code in the upcoming weeks…

051

Felix Koehler @felix-m-koehler.bsky.social · 07/01/2025

I created a video to help you get started using the APEBench suite (my recent #neurips paper) to benchmark autoregressive neural emulators for PDEs with a simple ConvNet emulation of 1D advection: youtu.be/q8fjQ4ZFynw

youtu.be

APEBench Quickstart

YouTube video by Machine Learning & Simulation

030

Felix Koehler @felix-m-koehler.bsky.social · 02/01/2025

Happy new year! 🎉 Two days ago we entered 2025 and just in time the channel surpassed 25k subscribers. Wow! Thanks to everyone for their kind words and support along the way: www.youtube.com/channel/UCh0...

A screenshot of the YouTube channel homepage with "25K subscribers" highlighted in a red hand-drawn circle.

080

Felix Koehler @felix-m-koehler.bsky.social · 20/12/2024

Check out my latest video on approximating the largest Lyapunov exponent of a dynamical system by integrating a tangent linear perturbation dynamic via autodiff in JAX: youtu.be/zRMBIkpcuu0 Very neat use-case of forward-mode AD for efficient Lyap approximation.

youtu.be

Largest Lyapunov Exponent using Autodiff in JAX/Python

YouTube video by Machine Learning & Simulation

010

Reposted by Felix Koehler

Samuel Vaiter @samuelvaiter.com · 13/12/2024

Automatic differentiation in forward mode computes derivatives by breaking down functions into elem operations and propagating derivatives alongside values. It’s efficient for functions with fewer inputs than outputs and for Jacobian-vect prod, using for instance dual numbers.

23810

Reposted by Felix Koehler

Miles Cranmer @milescranmer.bsky.social · 11/12/2024

Excited to be at NeurIPS this week! 🎉 I'm part of four exciting projects being presented: - The Well & Multimodal Universe: massive, curated scientific datasets - LaSR: LLM concept evolution for symbolic regression - MPP: 0th gen @polymathicai.bsky.social All posters Wed/Thu - stop by! 👋

0153

Reposted by Felix Koehler

Samuel Vaiter @samuelvaiter.com · 11/12/2024

I am at #NEURIPS and will present it today Wed 11 at 11:00am, West Ballroom A-D #6001

0112

Felix Koehler @felix-m-koehler.bsky.social · 11/12/2024

I will be presenting my poster on APEBench on Thursday from 11:00 to 14:00 PST at West Ballroom A-D #5407. This was done as part of my PhD with @thuereygroup.bsky.social in collaboration with my talented co-author, Simon Niedermayr, who is supervised by Rüdiger Westermann.

142