48:10

Hao Ni - Path development network for sequential data analysis

48:34

Simon Du - How Over-Parameterization Slows Down Gradient Descent

49:33

Micah Goldblum - Bridging the Gap between Deep Learning Theory and Practice

54:19

Lei Wu - Understanding the implicit bias of SGD: A dynamical stability perspective

51:24

Bobak Kiani: On the hardness of learning under symmetries

01:00:05

Hemant Tyagi - Dynamic ranking and translation synchronization

46:41

Nicolas Boulle - Elliptic PDE learning is provably data-efficient

49:21

Yuan Cao - Understanding Deep Learning Through Phenomena Discovery and Explanation

44:50

Mufan Li - Infinite-Depth Neural Networks as Depthwise Stochastic Processes

50:35

Aditya Varre - On the spectral bias of two-layer linear networks

49:54

Shuyang Ling - Neural collapse phenomenon for unconstrained feature model with imbalanced datasets

55:37

Marius Zeinhofer - Error Analysis and Optimization Methods for Scientific Machine Learning

58:30

Keaton Hamm - Manifold Learning in Wasserstein Space

39:07

Tan Nguyen - Transformers Meet Image Denoising: Mitigating Over-smoothing in Transformers

51:17

Lisa Kreusser - Unveiling the role of the Wasserstein Distance in Generative Modelling

49:20

Ting Lin - Universal Approximation and Expressive Power of Deep Neural Networks

51:29

Theo Bourdais - Computational Hypergraph Discovery, a Gaussian Process framework

53:25

Sebastian Goldt - Gaussian world is not enough: Analysing neural nets beyond Gaussian models of data

39:07

Tan Nguyen - Transformers Meet Image Denoising: Mitigating Over-smoothing in Transformers

45:37

Jakwang Kim - Understanding adversarial robustness via optimal transport perspective

46:59

Zhiqin Xu - Simple bias in deep learning

45:10

Matthieu Darcy - Kernel methods for operator learning

46:25

Liyuan Liu - Bridge Discrete Variables with Back-Propagation and Beyond

42:03

Ekaterina Rapinchuk - A Fast Graph-Based Classification Method with Applications to 3D Sensory Data

51:11

Christian Fiedler - Reproducing kernel Hilbert spaces in the mean field limit

01:00:35

Jean-Francois Aujol - FISTA is a geometrically optimized algorithm for strongly convex functions

32:29

Somdatta Goswami - Transfer Learning in Physics-Based Applications with Deep Neural Operators

46:41

Yaoqing Yang - Predicting & improving generalization by measuring loss landscapes & weight matrices

01:06:07

Bertrand Gauthier - Energy-driven sampling for PSD-matrix low-rank approximation

48:00

Ying Jin - Prediction-Assisted Screening and Discovery with Conformal p-values

53:00

Kevin Miller - Ensuring Exploration and Exploitation in Graph-Based Active Learning

53:40

Jose Gallego-Posada - Controlled Sparsity via Constrained Optimization

57:03

Kathryn Lindsey - Images and fibers of the realization map for feedforward ReLU neural networks

43:24

Anirbit Mukherjee - Provable Training of Neural Nets With One Layer of Activation

41:24

Deanna Needell - Using Algebraic Factorizations for Interpretable Learning

41:54

Daniel Cremers - Self-Supervised Learning for 3D Shape Analysis

56:39

Elisenda Grigsby - Functional dimension of ReLU Networks

42:06

Rebekka Burkholz - Pruning Deep Neural Networks for Lottery Tickets

44:12

Soufiane Hayou - Principled scaling of deep neural networks

53:36

Anna Little - Unbiasing Procedures for Scale-invariant Multi-reference Alignment

56:12

Francis Bach - Information theory through kernel methods

01:01:36

Simone Brugiapaglia - foundations of deep learning: from rating impossibility to existence theorems

41:26

Johannes Brandstetter - Towards a new generation of neural PDE surrogates

58:22

Marcus Hutter - Testing Independence of Exchangeable Random Variables

54:26

Leon Bungert - Uniform convergence rates for infinity Laplacian equations on graphs

01:03:53

Sophie Langer - Circumventing the curse of dimensionality with deep neural networks

01:04:35

Peter Richtarik - The Resolution of a Question Related to Local Training in Federated Learning

58:28

Gal Vardi - Implications of the implicit bias in neural networks

01:00:05

Denny Wu - High-dimensional asymptotics of feature learning in the early phase of NN training

57:48

Stephan Mandt - Compressing Variational Bayes: From neural data compression to video prediction

50:28

Gregory Schwartzman - SGD Through the Lens of Kolmogorov Complexity

46:50

Robin Walters - Symmetry in Neural Network Parameters and Non-Linearities

56:57

Tyrus Berry - Beyond Regression: Operators and Extrapolation in Machine Learning

46:42

Matthew Colbrook - Smale’s 18th Problem and the Barriers of Deep Learning

56:06

Hongyang Zhang - Understanding and improving generalization in multitask and transfer learning

50:21

Houman Owhadi - Computational Graph Completion

46:38

Loucas Pillaud-Vivien - the role of noise in non-convex machine learning dynamics

48:37

Alessandro Scagliotti - Deep Learning Approximation of Diffeomorphisms via Linear-Control Systems

01:01:35

Chris Budd and Simone Appella - R-Adaptivity, Deep Learning and Optimal Transport

56:09

Soheil Kolouri - Wasserstein Embeddings in the Deep Learning Era