#ePrint Fast and Private Computation of Cardinality of Set Intersection and Union: E De Cristofaro, P Gasti, G Tsudik ia.cr/2011/141

Computer vision is coming back into the forefront with Stable Diffusion. But if you're totally new to CV, you've gotta get started somewhere. No matter your skill level, here’s my favorite computer vision course. (And, of course, it’s 100% free from University of Michigan!)

Kyushu University released more than 1500 3D models of animals and plants using a photogrammetry technique, taking digital photos from multiple angles [database: buff.ly/3OmYi9N] [full paper: buff.ly/3UTBq40]

Interested in sparse training? Quick feature selection in case of high dimensional feature space, many samples, or noisy environments? Excited to present our new sparse training method for feature selection at #NeurIPS2022 (Wed, 2pm Hall J #341)🧵 Zahra Atashgahi,Mykola Pechenizkiy,Decebal

🎄Instead of asking whether tree structure should be baked into NNs, our new paper (arxiv.org/abs/2211.01288) asks if transformers already have a tendency to learn tree structured computations when trained on language, and if this structure is predictive of generalization! (1/n)

Invariant Lipschitz Bandits: A Side Observation Approach. (arXiv:2212.07524v1 [cs.LG]) ift.tt/3fIr1dk

The brain is organized along a complex hierarchy. How does the brain take advantage of different cell types and connection topologies to effectively distribute information across this hierarchy? Find out in our recent paper out now in Science Advances! science.org/doi/10.1126/sc…

🎉New paper! The survey of deep learning for mathematical reasoning (#DL4MATH) is now available. We've seen tremendous growth in this community since 2018, and this review covers the tasks, datasets, and methods from the past decade. Check it out now: arxiv.org/abs/2212.10535

I wrote #beginner level book teaching Deep Learning - its goal is to be the easiest intro possible In the book, each lesson builds a neural component *from scratch* in #NumPy Each *from scratch* toy code example is in the Github below #100DaysOfMLCode github.com/iamtrask/Grokk…

For mathematicians interested in unusual applications of topology: have a look at the Principles of Topological Psychology by Kurt Lewin (1936), where you can learn about Boundaries of Psychological Regions and The Dimensions of Life Space

Research codebases get really cluttered over time. What are the nanoGPTs (github.com/karpathy/nanoG… by Andrej Karpathy) of other models/tasks?

Can deep transformers be trained without skip connections nor normalisation layers? Our ICLR 2023 paper shows you how, using wide NN signal propagation ideas. We hope this can potentially pave the way to more efficient deep LLMs! (1/9) Paper: arxiv.org/abs/2302.10322

📉Learning rate decay is super effective and sometimes mysterious, but the simplest model of SGD on quadratic loss w/ noisy gradients almost perfectly predicts loss curves of transformers trained with Adam on real data, across schedules, model sizes, and token budgets. 1/4

[LG] One Token to Fool LLM-as-a-Judge Y Zhao, H Liu, D Yu, S.Y. Kung... [Tencent AI Lab] (2025) arxiv.org/abs/2507.08794

genuinely the coolest blog ever: ampcode.com/how-to-build-a… sending this to an agent to make your first agent is nuts

I implemented GRPO and DPO from scratch in vanilla Pytorch to unravel every piece of training details. Hope it could be helpful for those who care about the implementation details of the algorithms. 👉 github.com/mingyin0312/RL… #AI #RL #LLM