We propose an alternative to symbol-permutation data augmentations for structured problem-solving tasks: By introducing a dedicated symbol axis into the recurrent state tensors of RRMs and using an attention-based design, SE‑RRMs can inherently preserve symbol eq. when needed.
Really nice to see ConGLUDe bringing together structure‑ and ligand‑based training in a single contrastive geometric architecture. Love that it builds upon VN-EGNN, that very recently appeared in the Journal of Cheminformatics (doi.org/10.1186/s133...).
Unfortunately the original Tox21 Challenge dataset benchmark was redefined in other studies, which limits comparability of the performance of different prediction methods.
Now my colleagues have created a reproducible leaderboard hosted on Hugging Face.
It's happening again!!!
ML4Molecules workshop 2025.
within the #ELLIS Unconference, preceding #EurIPS.
More infos: moleculediscovery.github.io/workshop2025/
Happy to introduce 🔥LaM-SLidE🔥!
We show how trajectories of spatial dynamical systems can be modeled in latent space by
--> leveraging IDENTIFIERS.
📚Paper: arxiv.org/abs/2502.12128
💻Code: github.com/ml-jku/LaM-S...
📝Blog: ml-jku.github.io/LaM-SLidE/
1/n
Need to predict bioactivity 🧪 but only have limited data ❌?
Try our interactive app for prompting MHNfs — a state-of-the-art model for few-shot molecule–property prediction. No coding or training needed. 🚀
📄 Paper:
pubs.acs.org/doi/10.1021/...
🖥️ App:
huggingface.co/spaces/ml-jk...
The Machine Learning for Molecules workshop 2024 will take place THIS FRIDAY, December 6.
Tickets for in-person participation are "SOLD" OUT.
We still have a few free tickets for online/virtual participation!
Registration link here: moleculediscovery.github.io/workshop2024/
Thank you!
Hi Willie, could I join your list as well?