torchode solves batches of ODEs independently in parallel and works with #pytorch JIT.
Meet me at our poster on Friday at the DLDE workshop @ #neurips2022 , 13:10 UTC or 18:05 UTC.
Paper: arxiv.org/abs/2210.12375
Code: github.com/martenlienen/t…
w/Stephan Günnemann
Excited to release the code & the camera-ready version of our #NeurIPS2020 oral paper 'Fast and Flexible Temporal Point Processes with Triangular Maps' with Nicholas Gao, Marin Biloš and Stephan Günnemann.
Paper: arxiv.org/abs/2006.12631
Code: github.com/shchur/triangu…
Thrilled to present Hierarchical Smoothing at #NeurIPS 2023 : Novel robustness certificates against sparse adversarial attacks!
Joint work with fantastic collaborators Jan Schuchardt Aleksandar Bojchevski Stephan Günnemann!
Paper: arxiv.org/abs/2310.16221
Poster #715 Tuesday 5-7 pm #NeurIPS
'Robustness certificate' (for #MachineLearning models for graphs) -> provable guarantee that preds does not change under some perturbations.
How to give such guarantee? Ask Stephan Günnemann at #ecmlpkdd2020 .
If you are at #neurips2022 and are interested in the robustness of Graph Neural Networks then stop by our poster (#906) this afternoon (Poster Session 6).
Joint work with Felix Mujkanovic, Stephan Günnemann and Aleksandar Bojchevski
Happy to announce our new paper on Deterministic Uncertainty Methods TrustML-(Un)Limited #ICLR2023 !
We dissect how the design of training schemes, architecture, and prior can significantly impact feature collapse and uncertainty performance!
w/ C. Zhang and Stephan Günnemann
Interested how to leverage chemical priors for drug response prediction in single-cell data?
I will present chemCPA at #NeurIPS2022 , today 11am at Hall J #524
Paper: openreview.net/forum?id=vRrFV…
Code: github.com/theislab/chemC…
w\Simon Boehm,Niki Kilbertus,Stephan Günnemann, @Mohlotf,Fabian Theis
Are Defenses for Graph Neural Networks Robust?
As you might have guessed, they are at least not as robust as advertised.
This is also the title of our new paper accepted to #NeurIPS2022
Joint work with Felix Mujkanovic Stephan Günnemann Aleksandar Bojchevski
So far, adversarial training has not stood out as the go-to defense for graph neural networks w.r.t. perturbations of the graph structure. This has changed now with our NeurIPS 2023 paper!
Joint work with Lukas Gosch, Daniel Sturm, Bertrand Charpentier Daniel Zuegner @[email protected] Stephan Günnemann
(1/5)🧵
If you are at #ICLR2023 and interested in the robustness of GNNs, then visit us at our Tuesday poster session (#156, MH1-2-3-4) between 11:30 am and 1:30 pm :).
Paper: openreview.net/pdf?id=h1o7Ry9…
Video: iclr.cc/virtual/2023/p…
Joint work with Daniel Sturm, Simon Geisler, Stephan Günnemann
I am happy to announce that I recently started my PhD at TU München under the supervision of Prof. Stephan Günnemann.
I look forward to diving into this new professional and personal adventure!
I am thrilled to present Add and Thin: Diffusion for Temporal Point Processes next week at #NeurIPS2023 .
Joint work with my fantastic collaborators Marin Bilos, Oleksandr Shchur Marten Lienen Stephan Günnemann!
Paper and code are available here: cs.cit.tum.de/daml/add-thin
More in the 🧵.
Finally, #NeurIPS2021 is on and we are proud to present our paper on:
Whole Brain Vessel Graphs: A Dataset and Benchmark for #graph Learning and #neuroscience
With this work we hope to introduce real biological datasets to #ml research. Ali Max Erturk Stephan Günnemann menze_group
If you are interested in ML potentials and/or uncertainty estimation come and visit our presentation of 'Uncertainty Estimation for Molecules: Desiderata and Methods' this week at #ICML ! Tue, 11am, #415
Joint work with Nicholas Gao Bertrand Charpentier Amine Ketata Stephan Günnemann
1/2
Does graph contrastive learning truly improve adversarial robustness?
We answer this question in our work at #GLFrontiers @ #NeurIPS2023 .
Paper: arxiv.org/abs/2311.17853
Poster: 16 Dec 4:30pm Hall C2
Joint work with Zinuo Yi, Anna Starovoit, Rafiق, Simon Geisler, Stephan Günnemann
The Learning on Graphs Conference 2023 has concluded and it was a brilliantly executed experience! Many thanks to the incredible organising committee: Hannes Stärk Derek Lim Chaitanya K. Joshi Andreea Deac to name but a few! I particularly enjoyed the keynotes from Stephan Günnemann and Petar Veličković!
On May 14th, our co-director Gitta Kutyniok will talk about “Reliable AI: From Mathematical Foundations to Neuromorphic Computing” at the next
Women in AI & Robotics community meetup co-organized by relAI, also featuring Prof. Elke Wolf! Please sign up: eventbrite.de/e/women-in-ai-… Stephan Günnemann
📢Call for applications to our relAI MSc program! Apply by June 17, 2024 to receive cross-sectional training in AI and a full scholarship for your Master studies at TU München or Universität München.
🔗zuseschoolrelai.de/application/
DAAD News Gitta Kutyniok Stephan Günnemann Research in Germany - Initiative of the BMBF
