New paper on arXiv! And I think it's a good'un 😄 Meet the new Lattice Random Walk (LRW) discretisation for SDEs. It’s radically different from traditional methods like Euler-Maruyama (EM) in that each iteration can only move in discrete steps {-δₓ, 0, δₓ}.

Samuel Duffield, Maxwell Aifer, Denis Melanson, Zach Belateche, Patrick J. Coles Lattice Random Walk Discretisations of Stochastic Differential Equations https://arxiv.org/abs/2508.20883

Luca Ambrogioni The Information Dynamics of Generative Diffusion https://arxiv.org/abs/2508.19897

Great stuff: arxiv.org/abs/2508.18175

A random old one: "Kernels and Decision Trees" hackmd.io/@sp-monte-ca...

It’s a thing! www.microsoft.com/en-us/resear...

Struggling to control a system under noise and uncertainty? Willem Esterhuizen's new article dives into Model Predictive Control (MPC), a powerful feedback loop that uses a model to anticipate and correct system behavior. Learn how to handle hard constraints and disturbances.

Eliot Beyler (SIERRA), Francis Bach (SIERRA): Convergence of Deterministic and Stochastic Diffusion-Model Samplers: A Simple Analysis in Wasserstein Distance https://arxiv.org/abs/2508.03210 https://arxiv.org/pdf/2508.03210 https://arxiv.org/html/2508.03210

Francisco Daunas, I\~naki Esnaola, Samir M. Perlaza A Dual Optimization View to Empirical Risk Minimization with f-Divergence Regularization https://arxiv.org/abs/2508.03314

Sergio Calvo-Ordonez, Matthieu Meunier, Alvaro Cartea, Christoph Reisinger, Yarin Gal, Jose Miguel Hernandez-Lobato: Weighted Conditional Flow Matching https://arxiv.org/abs/2507.22270 https://arxiv.org/pdf/2507.22270 https://arxiv.org/html/2507.22270

Alan Jeffares, Liyuan Liu: An Introduction to Discrete Variational Autoencoders https://arxiv.org/abs/2505.10344 https://arxiv.org/pdf/2505.10344 https://arxiv.org/html/2505.10344

Tomorrow we discuss diffusion models for sampling unnormalized densities "Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching" arxiv.org/abs/2504.11713 Join us on zoom at 9am PT / 12pm ET / 6pm CEST: portal.valencelabs.com/starklyspeak...

Yaxin Ma, Benjamin Colburn, Jose C. Principe: A Simple and Effective Method for Uncertainty Quantification and OOD Detection https://arxiv.org/abs/2508.00754 https://arxiv.org/pdf/2508.00754 https://arxiv.org/html/2508.00754

A very nice article from @physrevx.bsky.social journals.aps.org/prx/abstract...

Another very nice paper arxiv.org/abs/2502.02300

I had a great time presenting "It's Time to Say Goodbye to Hard Constraints" at the Flatiron Institute. In this talk, I describe a philosophy for model construction in machine learning. Video now online! www.youtube.com/watch?v=LxuN...

Join Dr. Lianrui Zuo, Postdoctoral Researcher in ECE at VU and AI Scholar with VALIANT, at AI Summer School! 🌟 He'll be presenting "The Shape of Data in Noise: Diffusion Models as a Programmable Prior" Registration ends on July 31st. Register here: buff.ly/4fyVotP

Ensemble Kalman Diffusion Guidance: A Derivative-free Method for Inverse Problems Hongkai Zheng, Wenda Chu, Austin Wang et al. Action editor: Valentin De Bortoli https://openreview.net/forum?id=XPEEsKneKs #diffusion #kalman #inverse

Personalization of Large Language Models: A Survey Zhehao Zhang, Ryan A. Rossi, Branislav Kveton et al. Action editor: Sarath Chandar https://openreview.net/forum?id=tf6A9EYMo6 #personalization #personalized #formalization

New in the #DeeperLearningBlog: #KempnerInstitute researchers @binxuwang.bsky.social and John J. Vastola explain their work uncovering the linear Gaussian structure in diffusion models and the potential to use it to enhance performance. bit.ly/4lCauDv #AI #DiffusionModels

Join Dr. Daniel Moyer, Assistant Professor of Computer Science at Vanderbilt University, at AI Summer School! 🌟 He'll be presenting "Second Steps in Neural Networks" Space is limited. Registration ends on July 31st. Register here: buff.ly/x4pyQDo

Join Dr. Soheil Kolouri, Assistant Professor of CS and ECE at Vanderbilt University, at AI Summer School! 🌟 He'll be presenting "A Crash Course on Optimal Transport and Wasserstein Distances" Registration ends on July 31st. Register here: buff.ly/4fyVotP

Gianluigi Silvestri, Luca Ambrogioni CoVAE: Consistency Training of Variational Autoencoders https://arxiv.org/abs/2507.09103

Diffusion Model Predictive Control Guangyao Zhou, Sivaramakrishnan Swaminathan, Rajkumar Vasudeva Raju et al. Action editor: Stephen James https://openreview.net/forum?id=pvtgffHtJm #planning #reinforcement #learns

Pascanu, Lyle, Modoranu, Borras, Alistarh, Velickovic, Chandar, De, Martens: Optimizers Qualitatively Alter Solutions And We Should Leverage This https://arxiv.org/abs/2507.12224 https://arxiv.org/pdf/2507.12224 https://arxiv.org/html/2507.12224

New #Featured Certification: The Geometry of Phase Transitions in Diffusion Models: Tubular Neighbourhoods and Singularities Manato Yaguchi, Kotaro Sakamoto, Ryosuke Sakamoto et al. https://openreview.net/forum?id=ahVFKFLYk2 #diffusion #models #phase

Neta Shaul, Uriel Singer, Itai Gat, Yaron Lipman: Transition Matching: Scalable and Flexible Generative Modeling https://arxiv.org/abs/2506.23589 https://arxiv.org/pdf/2506.23589 https://arxiv.org/html/2506.23589

🌌🛰️🔭Want to explore universal visual features? Check out our interactive demo of concepts learned from our #ICML2025 paper "Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment". Come see our poster at 4pm on Tuesday in East Exhibition hall A-B, E-1208!

New #Featured Certification: What Should Embeddings Embed? Autoregressive Models Represent Latent Generating Distributions Liyi Zhang, Michael Y. Li, R. Thomas McCoy et al. https://openreview.net/forum?id=YyMACp98Kz #embeddings #embed #memorization

Accelerated Diffusion Models via Speculative Sampling, at #icml25 ! 16:30 Tuesday July 15 poster E-3012 arxiv.org/abs/2501.05370 @vdebortoli.bsky.social Galashov @arnauddoucet.bsky.social

(Implicit) Ensembles of Ensembles: Epistemic Uncertainty Collapse in Large Models Andreas Kirsch Action editor: Mark Coates https://openreview.net/forum?id=ON7dtdEHVQ #ensembles #ensemble #ensembling

Unsupervised Anomaly Detection through Mass Repulsing Optimal Transport Eduardo Fernandes Montesuma, EL HABAZI Adel, Fred Maurice NGOLE MBOULA Action editor: Bernhard Geiger https://openreview.net/forum?id=PPGJ3EvENv #anomaly #anomalies #detection

Information Theoretic Guarantees For Policy Alignment In Large Language Models Youssef Mroueh, Apoorva Nitsure Action editor: Jonathan Berant https://openreview.net/forum?id=Uz9J77Riul #reward #rewards #policies

🚨 New preprint! 🚨 Everyone loves causal interp. It’s coherently defined! It makes testable predictions about mechanistic interventions! But what if we had a different objective: predicting model behavior not under mechanistic interventions, but on unseen input data?

The materials of my graduate school course at #SGP2025 on "Deep Learning on Meshes and Point Clouds" are live on rubenwiersma.nl/deeplearning. A 'map' for 3D deep learning, covering the basics and thoughts on when and where to use deep learning tools in 3D.

Augmented Invertible Koopman Autoencoder for long-term time series forecasting Anthony Frion, Lucas Drumetz, Mauro Dalla Mura, Guillaume Tochon, Abdeldjalil AISSA EL BEY Action editor: William Redman https://openreview.net/forum?id=o6ukhJLzMQ #autoencoders #encoder #autoencoder

New blog post: "Diffusion Elites: surprisingly good, simple and embarrassingly parallel", blog.christianperone.com/2025/07/diff...

Two fresh tutorials on low-dimensional DL: Sparsity ’23 🔗 vita-group.github.io/SparsityTuto... Low-Rank ’25 🔗 vita-group.github.io/LowrankTutor... Curious about how sparsity and low-rank ideas power modern deep learning? I hope these slide decks & notes are a useful starting point! 🚀

Accelerating Learned Image Compression Through Modeling Neural Training Dynamics Yichi Zhang, Zhihao Duan, Yuning Huang, Fengqing Zhu Action editor: Chang Xu https://openreview.net/forum?id=nannw4SGfS #compression #embedding #learned

Deep Augmentation: Dropout as Augmentation for Self-Supervised Learning Rickard Brüel Gabrielsson, Tongzhou Wang, Manel Baradad, Justin Solomon Action editor: Han Bao https://openreview.net/forum?id=OjWB2671AR #augmentation #augmentations #deep

M. M. A. Valiuddin, R. J. G. van Sloun, C. G. A. Viviers, P. H. N. de With, F. van der Sommen A Review of Bayesian Uncertainty Quantification in Deep Probabilistic Image Segmentation https://arxiv.org/abs/2411.16370

Machine Learning with Physics Knowledge for Prediction: A Survey Joe Watson, Chen Song, Oliver Weeger et al. Action editor: Markus Heinonen https://openreview.net/forum?id=ZiJYahyXLU #prediction #predictive #forecast

Hitchhiker's guide on the relation of Energy-Based Models with other generative models, sampling ... Davide Carbone Action editor: Alberto Bietti https://openreview.net/forum?id=VTgixSbrJI #generative #gans #autoencoders

The most important aspect when facing data shift is the type of shift present in the data. I will give below a few examples of shifts and some existing methods to compensate for it.🧵1/6

New Preprint: "On the Ability of Deep Networks to Learn Symmetries from Data: A Neural Kernel Theory" by Andrea Perin and myself. Retweets of this post or the thread below are highly appreciated! Follow @zazzarazzaz.bsky.social for future updates on this line of work.

Get ready for the PDE-Transformer: our new NN architecture tailored to scientific tasks 😁 It combines hierarchical processing (UDiT), scalability (SWin) and flexible conditioning mechanisms. Code and paper available at tum-pbs.github.io/pde-transfor...

Uniform Noise Distribution and Compact Clusters: Unveiling the Success of Self-Supervised Learnin... Pengcheng Xu, Li Yi, Gezheng Xu, Xi Chen, Ian McLeod, Charles Ling, Boyu Wang Action editor: Gang Niu https://openreview.net/forum?id=LDBjgS5Ez7 #supervised #labels #classifier

Lecture 14: Measurement-only dynamics: monitored transverse-field Ising model, entanglement phase transition & percolation, scrambling and unscrambling by measurement, monitored Kitaev model vimeo.com/1051163805

New #Survey Certification: A Survey of Reinforcement Learning from Human Feedback Timo Kaufmann, Paul Weng, Viktor Bengs, Eyke Hüllermeier https://openreview.net/forum?id=f7OkIurx4b #reinforcement #learns #feedback

Very simple idea, well executed arxiv.org/abs/2501.19195