Samir Moustafa @SamirMoustafa97

Research and Teaching Assistant at @univienna samirmoustafa.net Vienna, Austria Joined August 2020

Tweets

14
Followers

33
Following

222
Likes

52

Samir Moustafa @SamirMoustafa97

3 weeks ago

Excited to share our latest project, Torch Floating Point! Ever wondered how computers handle decimal numbers? We've built a tool that lets you create custom number formats and seamlessly integrate them with @PyTorch Check it out: github.com/SamirMoustafa/…

0 0 1 17 0

Download Image

Samir Moustafa @SamirMoustafa97

2 months ago

A few days ago, I received a Certificate of Appreciation from the @csunivie for my 2024 research contributions. Thanks for recognizing the chaos, coffee, and collaboration behind every paper. To W.Gansterer, @nmkriege , collaborators, and my incredible squad: This is our win.

0 0 4 78 0

Download Image

Samir Moustafa @SamirMoustafa97

a year ago

Our paper "Attacking Graph Neural Networks with Bit Flips: Weisfeiler and Leman Go Indifferent" was accepted for KDD 2024! Thanks to Lorenz Kummer, @sschritt, @nmkriege, and Wilfried Gansterer. See our work here: arxiv.org/abs/2311.01205 #KDD2024 #GraphNeuralNetworks #GNN

0 1 9 589 0

Nils M. Kriege @nmkriege

3 years ago

Open PhD student position at @csunivie, @univienna in the project "StruDL: Structured Data Learning with General Similarities" funded by @WWTF. See docs.univie.ac.at/fileadmin/user… for details, apply via docs.univie.ac.at/how-to-apply/p… with reference number "#03 PC4-23" by April 4, 2023.

1 4 7 803 1

Alex Tamkin @AlexTamkin

4 years ago

image-scaling attacks are wild small dots added to the image on the left turns it into the image on the right when downscaled could make auditing ML systems very tricky if you only look at the original images...

33 502 2K 0 303

Download Image

Chaitanya K. Joshi @chaitjo

6 years ago

Excited to share a blog post on the connection between #Transformers for NLP and #GraphNeuralNetworks (GNNs or GCNs). graphdeeplearning.github.io/post/transform…

16 240 825 0 283

Download Image

Andrej Karpathy @karpathy

5 years ago

I wrote a minimal/educational GPT training library in PyTorch, am calling it minGPT as it is only around ~300 lines of code: github.com/karpathy/minGPT +demos for addition and character-level language model. (quick weekend project, may contain sharp edges)

35 631 3K 0 560

PyTorch @PyTorch

5 years ago

PyTorch Lightning V0.9 is available now featuring the final API with better data decoupling, shorter logging syntax, synchronized batchnorm, and more. Learn more: medium.com/pytorch/pytorc…

4 117 498 0 29

Google AI @GoogleAI

5 years ago

To address the challenges unique to offline #ReinforcementLearning, we are releasing an open-source benchmark, D4RL, as well as a simple and effective offline RL algorithm, called conservative Q-learning (CQL). Read all about it at goo.gle/34vbt3J

1 147 390 0 50

Download Image

Rod Rivera @rodriveracom

5 years ago

Recently, I gave a lesson on design patterns for Python. The bad thing is that few students of Data Science know about them. The good thing is that Python is so well-designed that one does not need most of them, one just needs PEP-8 and to write Pythonesque code.