AMLab
amlab.bsky.social
The official account of the Amsterdam Machine Learning Lab (AMLab) at UvA, co-directed by Max Welling and Jan-Willem van de Meent.
👤 Joined on 17 November 2024
809
Followers
40
Following
43
Total Posts
📜 Kullanıcı Geçmişi
- 17 Kas 2024, 10:33
🎉 Bluesky'e katıldı - İlk kullanıcı adı:
@amlab.bsky.social
🚫 Blokladığı Kullanıcılar
❤️ Beğendiği Gönderiler
thats a wrap! you can find the code here: github.com/maxxxzdn/erwin and here is a cute visualisation of Ball Tree building:
Show PostTo scale Erwin further to gigantic industrial datasets, we combine Transolver and Erwin by applying ball tree attention over latent tokens. The key insight is that by using Erwin, we can afford larger bottleneck sizes while maintaining efficiency. Stay tuned for updates!
Show PostTo improve the processing of long-range information while keeping the cost sub-quadratic, we combine the ball tree with Native Sparse Attention (NSA), which align naturally with each other. Accepted to Long-Context Foundation Models at ICML 2025! paper: arxiv.org/abs/2506.12541
Show PostThere are many possible directions for building upon this work. I will highlight two on which I have been working together with my students.
Show PostOn top of the initial experiments in the first version of the paper, we include two more PDE-related benchmarks, where Erwin shows strong performance, achieving state-of-the-art on multiple tasks.
Show Post