01.06.2023, 17:00 Uhr:

• Rishi Sonthalia (UCLA)
• Least Squares Denoising: Non-IID Data, Transfer Learning and Under-parameterized Double Descent
• Live Stream, registration required, check abstract for details

08.06.2023, 17:00 Uhr:

• Simone Bombari (IST Austria)
• to be announced
• Live Stream, registration required, check abstract for details

06.07.2023, 17:00 Uhr:

• Ronen Basri (Weizmann Institute of Science)
• to be announced
• Live Stream, registration required, check abstract for details

17.08.2023, 17:00 Uhr:

• Surbhi Goel (University of Pennsylvania)
• to be announced
• Live Stream, registration required, check abstract for details

02.04.2020, 11:00 Uhr:

• Yu Guang Wang (University of New South Wales, Sydney)
• Haar Graph Pooling
• see Abstract
• see video

09.04.2020, 17:00 Uhr:

• Michael Arbel (University College London)
• Kernelized Wasserstein Natural Gradient
• see video

16.04.2020, 17:00 Uhr:

• Anton Mallasto (University of Copenhagen)
• Estimation of Optimal Transport in Generative Models
• see video

23.04.2020, 17:00 Uhr:

• Johannes Mueller (MPI MiS, Leipzig)
• Deep Ritz Revisited
• see video

30.04.2020, 17:00 Uhr:

• Quynh Nguyen (University Saarbruecken)
• Loss surface of deep and wide neural networks
• see video

07.05.2020, 17:00 Uhr:

• Kathlén Kohn (KTH Royal Institute of Technology)
• The geometry of neural networks
• see video

14.05.2020, 17:00 Uhr:

• Benjamin Fehrman (University of Oxford)
• Convergence rates for the stochastic gradient descent method for non-convex objective functions
• see video

21.05.2020, 17:00 Uhr:

• Dennis Elbrächter (Universität Wien)
• How degenerate is the parametrization of (ReLU) neural networks?
• see video

28.05.2020, 17:00 Uhr:

• Jonathan Frankle (Massachusetts Institute of Technology)
• The Lottery Ticket Hypothesis: On Sparse, Trainable Neural Networks

04.06.2020, 17:00 Uhr:

• Ulrich Terstiege (Rheinisch-Westfälische Technische Hochschule Aachen)
• Learning deep linear neural networks: Riemannian gradient flows and convergence to global minimizers

11.06.2020, 17:00 Uhr:

• Poorya Mianjy (Johns Hopkins University)
• Understanding the Algorithmic Regularization due to Dropout
• see video

18.06.2020, 17:00 Uhr:

• Alessandro Achille (University of California, Los Angeles)
• Structure of Learning Tasks and the Information in the Weights of a Deep Network
• see video

25.06.2020, 17:00 Uhr:

• Adam Gaier (INRIA)
• Weight Agnostic Neural Networks

02.07.2020, 17:00 Uhr:

• Wenda Zhou (Columbia University)
• New perspectives on cross-validation
• see video

03.07.2020, 17:00 Uhr:

• Kai Fong Ernest Chong (Singapore University of Technology and Design)
• The approximation capabilities of neural networks
• see video

10.07.2020, 17:00 Uhr:

• Nasim Rahaman (MPI-IS Tübingen, and Mila, Montréal)
• On the Spectral Bias of Neural Networks
• see video

23.07.2020, 17:00 Uhr:

• Léonard Blier (Facebook AI Research, Université Paris Saclay, Inria)
• The Description Length of Deep Learning Models

30.07.2020, 17:00 Uhr:

• Robert Peharz (TU Eindhoven)
• Minimal Random Code Learning: Getting Bits Back from Compressed Model Parameters
• see video

31.07.2020, 17:00 Uhr:

• Greg Ongie (University of Chicago)
• A function space view of overparameterized neural networks
• see video

06.08.2020, 17:00 Uhr:

• Simon S. Du (University of Washington)
• Ultra-wide Neural Network and Neural Tangent Kernel
• see video

13.08.2020, 17:00 Uhr:

• Lénaïc Chizat (CNRS - Laboratoire de Mathématiques d'Orsay)
• Analysis of Gradient Descent on Wide Two-Layer ReLU Neural Networks
• see video

14.08.2020, 17:00 Uhr:

• Ido Nachum (École Polytechnique Fédérale de Lausanne)
• Regularization by Misclassification in ReLU Neural Networks
• see video

20.08.2020, 17:00 Uhr:

• Arthur Jacot (École Polytechnique Fédérale de Lausanne)
• Neural Tangent Kernel: Convergence and Generalization of DNNs
• see video

21.08.2020, 17:00 Uhr:

• Preetum Nakkiran (Harvard University)
• Distributional Generalization: A New Kind of Generalization
• see video

27.08.2020, 17:00 Uhr:

• Felix Draxler (Heidelberg University)
• Characterizing The Role of A Single Coupling Layer in Affine Normalizing Flows
• see video

03.09.2020, 17:00 Uhr:

• Niladri S. Chatterji (UC Berkeley)
• Upper and lower bounds for gradient based sampling methods
• see video

04.09.2020, 17:00 Uhr:

• Nadav Cohen (Tel Aviv University)
• Analyzing Optimization and Generalization in Deep Learning via Dynamics of Gradient Descent
• see video

17.09.2020, 17:00 Uhr:

• Mahito Sugiyama (National Institute of Informatics, JST, PRESTO)
• Learning with Dually Flat Structure and Incidence Algebra

18.09.2020, 17:00 Uhr:

• Mert Pilanci (Stanford University)
• Exact Polynomial-time Convex Formulations for Training Multilayer Neural Networks: The Hidden Convex Optimization Landscape

24.09.2020, 17:00 Uhr:

• Liwen Zhang (University of Chicago)
• Tropical Geometry of Deep Neural Networks
• see video

25.09.2020, 17:00 Uhr:

• Randall Balestriero (Rice University)
• Max-Affine Spline Insights into Deep Networks
• see video

01.10.2020, 17:00 Uhr:

• David Rolnick (McGill University & Mila)
• Expressivity and learnability: linear regions in deep ReLU networks
• see video

08.10.2020, 17:00 Uhr:

• Yaoyu Zhang (Shanghai Jiao Tong University)
• Impact of initialization on generalization of deep neural networks
• see video

15.10.2020, 17:00 Uhr:

• Guy Blanc (Stanford University)
• Provable Guarantees for Decision Tree Induction
• see video

22.10.2020, 17:00 Uhr:

• Boris Hanin (Princeton University)
• Neural Networks at Finite Width and Large Depth
• see video

29.10.2020, 17:00 Uhr:

• Yaim Cooper (Institute for Advanced Study, Princeton)
• The geometry of the loss function of deep neural networks

05.11.2020, 17:00 Uhr:

• Kenji Kawaguchi (MIT)
• Deep learning: theoretical results on optimization and mixup

12.11.2020, 17:00 Uhr:

• Yasaman Bahri (Google Brain)
• The Large Learning Rate Phase of Wide, Deep Neural Networks
• see video

20.11.2020, 17:00 Uhr:

• Amartya Sanyal (University of Oxford)
• How benign is benign overfitting?
• see video

17.12.2020, 17:00 Uhr:

• Tim G. J. Rudner (University of Oxford)
• Outcome-Driven Reinforcement Learning via Variational Inference

14.01.2021, 17:00 Uhr:

• Mahdi Soltanolkotabi (University of Southern California)
• Learning via early stopping and untrained neural nets
• see video

21.01.2021, 17:00 Uhr:

• Taiji Suzuki (The University of Tokyo, and Center for Advanced Intelligence Project, RIKEN, Tokyo)
• Statistical efficiency and optimization of deep learning from the viewpoint of non-convexity
• see video

28.01.2021, 17:00 Uhr:

• Suriya Gunasekar (Microsoft Research, Redmond)
• Rethinking the role of optimization in learning
• see video

04.02.2021, 17:00 Uhr:

• Tailin Wu (Stanford University)
• Phase transitions on the universal tradeoff between prediction and compression in machine learning
• see video

11.02.2021, 17:00 Uhr:

• Samuel L. Smith (DeepMind)
• A Backward Error Analysis for Random Shuffling SGD
• see video

18.02.2021, 17:00 Uhr:

• Umut Şimşekli (INRIA - École Normale Supérieure (Paris))
• Towards Building a Heavy-Tailed Theory of Stochastic Gradient Descent for Deep Neural Networks
• see video

25.02.2021, 17:00 Uhr:

• Zhiyuan Li (Princeton University)
• Reconciling Modern Deep Learning with Traditional Optimization Analyses: The Intrinsic Learning Rate
• see video

04.03.2021, 17:00 Uhr:

• Marcus Hutter (DeepMind and Australian National University)
• Learning Curve Theory
• see video

11.03.2021, 17:00 Uhr:

• Spencer Frei (Department of Statistics, UCLA)
• Generalization of SGD-trained neural networks of any width in the presence of adversarial label noise
• see video

18.03.2021, 17:00 Uhr:

• Ryo Karakida (AIST (National Institute of Advanced Industrial Science and Technology), Tokyo)
• Understanding Approximate Fisher Information for Fast Convergence of Natural Gradient Descent in Wide Neural Networks
• see video

25.03.2021, 17:00 Uhr:

• Pratik Chaudhari (University of Pennsylvania)
• Learning with few labeled data
• see video

01.04.2021, 17:00 Uhr:

• Roi Livni (Tel Aviv University)
• Regularization, what is it good for?
• see video

08.04.2021, 17:00 Uhr:

• Cristian Bodnar (University of Cambridge)
• Fabrizio Frasca (Twitter)
• Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks
• see video

15.04.2021, 17:00 Uhr:

• Stanislav Fort (Stanford University)
• Neural Network Loss Landscapes in High Dimensions: Theory Meets Practice

22.04.2021, 17:00 Uhr:

• Matus Jan Telgarsky (UIUC)
• Recent advances in the analysis of the implicit bias of gradient descent on deep networks
• see video

29.04.2021, 17:00 Uhr:

• Mehrdad Farajtabar (DeepMind)
• Catastrophic Forgetting in Continual Learning of Neural Networks
• see video

06.05.2021, 17:00 Uhr:

• Haizhao Yang (Purdue University)
• A Few Thoughts on Deep Learning-Based PDE Solvers
• see video

13.05.2021, 17:00 Uhr:

• Eirikur Agustsson (Google research)
• Universally Quantized Neural Compression

27.05.2021, 17:00 Uhr:

• Sebastian Reich (University of Potsdam and University of Reading)
• Statistical inverse problems and affine-invariant gradient flow structures in the space of probability measures
• see slides

03.06.2021, 17:00 Uhr:

• Stanisław Jastrzębski (Molecule.one and Jagiellonian University)
• Reverse-engineering implicit regularization due to large learning rates in deep learning
• see video

10.06.2021, 17:00 Uhr:

• Isabel Valera (Saarland University and MPI for Intelligent Systems)
• Algorithmic recourse: theory and practice
• see video

17.06.2021, 17:00 Uhr:

• Ziv Goldfeld (Cornell University)
• Scaling Wasserstein distances to high dimensions via smoothing
• see video

01.07.2021, 17:00 Uhr:

• Dominik Janzing (Amazon Research, Tuebingen, Germany)
• Why we should prefer simple causal models

15.07.2021, 17:00 Uhr:

• Hossein Mobahi (Google Research)
• Self-Distillation Amplifies Regularization in Hilbert Space
• see video

22.07.2021, 17:00 Uhr:

• Soledad Villar (Johns Hopkins University)
• Scalars are universal: Gauge-equivariant machine learning, structured like classical physics
• see video

26.08.2021, 17:00 Uhr:

• Guodong Zhang (University of Toronto)
• Differentiable Game Dynamics: Hardness and Complexity of Equilibrium Learning
• see video

02.09.2021, 17:00 Uhr:

• Matthew Trager (Amazon US)
• A Geometric View of Functional Spaces of Neural Networks

09.09.2021, 17:00 Uhr:

• Shaowei Lin (working in a stealth startup)
• All you need is relative information
• see video

16.09.2021, 17:00 Uhr:

• Pavel Izmailov (New York University)
• What Are Bayesian Neural Network Posteriors Really Like?
• see video

23.09.2021, 17:00 Uhr:

• Huan Xiong (MBZUAI, Abu Dhabi)
• On the Number of Linear Regions of Convolutional Neural Networks with Piecewise Linear Activations
• see video

07.10.2021, 17:00 Uhr:

• Roger Grosse (University of Toronto)
• Self-tuning networks: Amortizing the hypergradient computation for hyperparameter optimization
• see video

14.10.2021, 17:00 Uhr:

• Mathias Drton (Technical University of Munich)
• A Bayesian information criterion for singular models
• see video

21.10.2021, 17:00 Uhr:

• David Stutz (MPI for Informatics, Saarbrücken)
• Relating Adversarial Robustness and Weight Robustness Through Flatness
• see video

04.11.2021, 17:00 Uhr:

• Hyeyoung Park (Kyungpook National University)
• Effect of Geometrical Singularity on Learning Dynamics of Neural Networks
• see video

11.11.2021, 17:00 Uhr:

• Christoph Hertrich (TU Berlin)
• Towards Lower Bounds on the Depth of ReLU Neural Networks
• see video

18.11.2021, 17:00 Uhr:

• Daniel Soudry (Technion)
• Algorithmic Bias Control in Deep learning
• see video

25.11.2021, 17:00 Uhr:

• Michael Joswig (TU Berlin)
• Geometric Disentanglement by Random Convex Polytopes
• see video

02.12.2021, 17:00 Uhr:

• Ekaterina Lobacheva (HSE University)
• Maxim Kodryan (HSE University)
• On the Periodic Behavior of Neural Network Training with Batch Normalization and Weight Decay
• see video

16.12.2021, 17:00 Uhr:

• Zhi-Qin John Xu (Shanghai Jiao Tong University)
• Occam’s razors in neural networks: frequency principle in training and embedding principle of loss landscape
• see video

13.01.2022, 17:00 Uhr:

• Yuan Cao (University of California, LA)
• Towards Understanding and Advancing the Generalization of Adam in Deep Learning
• see video

20.01.2022, 17:00 Uhr:

• Philipp Petersen (University of Vienna)
• Optimal learning in high-dimensional classification problems
• see video

27.01.2022, 17:00 Uhr:

• Daniel Russo (Columbia University)
• Adaptivity and Confounding in Multi-armed Bandit Experiments
• see video

04.02.2022, 17:00 Uhr:

• Song Mei (University of California, Berkeley)
• The efficiency of kernel methods on structured datasets
• see video

10.02.2022, 17:00 Uhr:

• Hui Jin (University of California, Los Angeles)
• Learning curves for Gaussian process regression with power-law priors and targets
• see video

17.02.2022, 17:00 Uhr:

• Maksim Velikanov (Skolkovo Institute of Science and Technology)
• Spectral properties of wide neural networks and their implications to the convergence speed of different Gradient Descent algorithms

24.02.2022, 17:00 Uhr:

• Danijar Hafner (Google Brain & University of Toronto)
• General Infomax Agents through World Models
• see video
• see slides

03.03.2022, 17:00 Uhr:

• Marco Mondelli (Institute of Science and Technology Austria)
• Gradient Descent for Deep Neural Networks: New Perspectives from Mean-field and NTK
• see video
• see slides

04.03.2022, 17:00 Uhr:

• Daniel McKenzie (UCLA)
• Implicit Neural Networks: What they are and how to use them.
• see video

10.03.2022, 17:00 Uhr:

• Gergely Neu (Universitat Pompeu Fabra, Barcelona)
• Generalization Bounds via Convex Analysis
• see video

11.03.2022, 17:00 Uhr:

• Uri Alon (CMU)
• Bottlenecks, over-squashing, and attention in Graph Neural Networks.
• see video

17.03.2022, 17:00 Uhr:

• Gabin Maxime Nguegnang (RWTH Aachen University)
• Convergence of gradient descent for learning deep linear neural networks
• see video

24.03.2022, 17:00 Uhr:

• Ohad Shamir (Weizmann Institute of Science)
• Implicit bias in machine learning
• see video
• see slides

31.03.2022, 17:00 Uhr:

• Benjamin Eysenbach (Carnegie Mellon University & Google Brain)
• The Information Geometry of Unsupervised Reinforcement Learning
• see video

07.04.2022, 17:00 Uhr:

• Chaoyue Liu (Ohio State University)
• Transition to Linearity of Wide Neural Networks
• see video

14.04.2022, 17:00 Uhr:

• Christoph Lampert (Institute of Science and Technology, Austria)
• Robust and Fair Multisource Learning
• see video

21.04.2022, 17:00 Uhr:

• Berfin Şimşek (Ecole Polytechnique Fédérale de Lausanne (EPFL))
• The Geometry of the Loss Landscape in Overparameterized Neural Networks: Symmetry-Induced Saddles and Global Minima Manifold
• see video
• see slides

05.05.2022, 17:00 Uhr:

• Noam Razin (Tel Aviv University)
• Generalization in Deep Learning Through the Lens of Implicit Rank Minimization
• see slides

12.05.2022, 17:00 Uhr:

• Zixiang Chen (UCLA)
• Benign Overfitting in Two-layer Convolutional Neural Networks
• see video
• see slides

19.05.2022, 17:00 Uhr:

• Sayan Mukherjee (MPI MiS, Leipzig + Universität Leipzig)
• Inference in dynamical systems and the geometry of learning group actions
• see video

26.05.2022, 17:00 Uhr:

• Melikasadat Emami (UCLA)
• Asymptotics of Learning in Generalized Linear Models and Recurrent Neural Networks
• see video

02.06.2022, 17:00 Uhr:

• Tan Nguyen (UCLA)
• Principled Models for Machine Learning
• see video

09.06.2022, 17:00 Uhr:

• Guang Cheng (UCLA)
• Nonparametric Perspective on Deep Learning
• see video

16.06.2022, 17:00 Uhr:

• Jingfeng Wu (Johns Hopkins University)
• A Fine-Grained Characterization for the Implicit Bias of SGD in Least Square Problems
• see video
• see slides

07.07.2022, 17:00 Uhr:

• Sebastian Kassing (University of Bielefeld)
• Convergence of Stochastic Gradient Descent for analytic target functions
• see video
• see slides

14.07.2022, 17:00 Uhr:

• Rong Ge (Duke University)
• Towards Understanding Training Dynamics for Mildly Overparametrized Models
• see video

21.07.2022, 17:00 Uhr:

• Mark Schmidt (University of British Columbia)
• Optimization Algorithms for Training Over-Parameterized Models
• see video
• see slides

28.07.2022, 17:00 Uhr:

• Arash A. Amini (UCLA)
• Target alignment in truncated kernel ridge regression
• see video

04.08.2022, 17:00 Uhr:

• Wei Hu (University of California, Berkeley)
• More Than a Toy: Random Matrix Models Predict How Real-World Neural Representations Generalize
• see video
• see slides

12.08.2022, 17:00 Uhr:

• Francesco Di Giovanni (Twitter)
• Over-squashing and over-smoothing through the lenses of curvature and multi-particle dynamics
• see video
• see slides

18.08.2022, 17:00 Uhr:

• Aditya Golatkar (UCLA)
• Selective Forgetting and Mixed Privacy in Deep Networks
• see video

25.08.2022, 17:00 Uhr:

• Melanie Weber (Harvard University)
• Exploiting geometric structure in (matrix-valued) optimization
• see video

08.09.2022, 17:00 Uhr:

• Devansh Arpit (Salesforce)
• Optimization Aspects that Improve IID and OOD Generalization in Deep Learning
• see video
• see slides

15.09.2022, 17:00 Uhr:

• Lisa Maria Kreusser (University of Bath)
• Wasserstein GANs Work Because They Fail (to Approximate the Wasserstein Distance)
• see slides

22.09.2022, 17:00 Uhr:

• Gintare Karolina Dziugaite (Google Brain)
• Deep Learning through the Lens of Data
• see video

06.10.2022, 17:00 Uhr:

• Felix Dietrich (Technical University of Munich)
• Learning dynamical systems from data
• see video

13.10.2022, 17:00 Uhr:

• Chao Ma (Stanford University)
• Implicit bias of optimization algorithms for neural networks: static and dynamic perspectives
• see video
• see slides

20.10.2022, 17:00 Uhr:

• Joe Kileel (University of Texas at Austin)
• The Effect of Parametrization on Nonconvex Optimization Landscapes
• see video

27.10.2022, 17:00 Uhr:

• Lechao Xiao (Google Brain)
• Eigenspace restructuring: A Principle of space and frequency in neural networks
• see video

03.11.2022, 17:00 Uhr:

• Itay Safran (Purdue University)
• On the Effective Number of Linear Regions in Shallow Univariate ReLU Networks: Convergence Guarantees and Implicit Bias
• see video

10.11.2022, 17:00 Uhr:

• El Mehdi Achour (IMT Toulouse)
• The loss landscape of deep linear neural networks: A second-order analysis
• see video
• see slides

17.11.2022, 17:00 Uhr:

• Sebastian Goldt (SISSA Trieste)
• On the importance of higher-order input statistics for learning in neural networks

01.12.2022, 17:00 Uhr:

• Levon Nurbekyan (UCLA)
• Efficient natural gradient method for large-scale optimization problems
• see video

08.12.2022, 17:00 Uhr:

• Mojtaba Sahraee-Ardakan (UCLA)
• Equivalence of Kernel Methods and Linear Models in High Dimensions
• see video

15.12.2022, 17:00 Uhr:

• Nicolás García Trillos (University of Wisconsin-Madison)
• Analysis of adversarial robustness and of other problems in modern machine learning
• see video
• see slides

16.12.2022, 17:00 Uhr:

• Jona Lelmi (University of Bonn)
• Large data limit of the MBO scheme for data clustering
• see video
• see slides

12.01.2023, 17:00 Uhr:

• Felipe Suárez-Colmenares (MIT)
• Learning threshold neurons via the "edge of stability”
• see video
• see slides

19.01.2023, 17:00 Uhr:

• Xiaowu Dai (UCLA)
• Kernel Ordinary Differential Equations
• see video

26.01.2023, 17:00 Uhr:

• Liu Ziyin (University of Tokyo)
• The Probabilistic Stability and Low-Rank Bias of SGD
• see video

02.02.2023, 17:00 Uhr:

• Mihaela Rosca (DeepMind)
• On continuous time models of gradient descent and instability in deep learning
• see video

09.02.2023, 17:00 Uhr:

• Renjie Liao (University of British Columbia)
• Gaussian-Bernoulli RBMs Without Tears
• see video

16.02.2023, 17:00 Uhr:

• Julia Elisenda Grigsby (Boston College)
• Functional dimension of ReLU Networks
• see video

23.02.2023, 17:00 Uhr:

• Jalaj Bhandari (Columbia University)
• Global guarantees for policy gradient methods.
• see video
• see slides

02.03.2023, 17:00 Uhr:

• Samy Wu Fung (Colorado School of Mines)
• Using Hamilton Jacobi PDEs in Optimization
• see video
• see slides

09.03.2023, 17:00 Uhr:

• Ricardo S. Baptista (Caltech)
• On the representation and learning of triangular transport maps
• see video
• see slides

16.03.2023, 17:00 Uhr:

• Chulhee Yun (KAIST)
• Shuffling-based stochastic optimization methods: bridging the theory-practice gap
• see video

23.03.2023, 17:00 Uhr:

• Dohyun Kwon (University of Wisconsin-Madison)
• Convergence of score-based generative modeling
• see video

06.04.2023, 17:00 Uhr:

• Jaehoon Lee (Google Brain)
• Exploring Infinite-Width Limit of Deep Neural Networks
• see video

13.04.2023, 17:00 Uhr:

• Wei Zhu (University of Massachusetts Amherst)
• Implicit Bias of Linear Equivariant Steerable Networks
• see video
• see slides

20.04.2023, 17:00 Uhr:

• Aditya Grover (UCLA)
• Generative Decision Making Under Uncertainty
• see video

27.04.2023, 17:00 Uhr:

• Stéphane d'Ascoli (ENS and FAIR Paris)
• Double descent: insights from the random feature model
• see video

11.05.2023, 17:00 Uhr:

• Henning Petzka (RWTH Aachen University)
• Relative flatness and generalization
• see video

18.05.2023, 17:00 Uhr:

• Duc N.M Hoang (University of Texas, Austin)
• Revisiting Pruning as Initialization through the Lens of Ramanujan Graph
• see video

25.05.2023, 17:00 Uhr:

• Łukasz Dębowski (IPI PAN)
• A Simplistic Model of Neural Scaling Laws: Multiperiodic Santa Fe Processes
• Link
• see slides