Workshops

Summer 2019

Deep Learning Reunion

Aug 10, 2020 to Aug 13, 2020

Return to event »

Monday, August 10th, 2020
9:50 am–10:00 am		Opening Remarks
10:00 am–10:15 am		Benign Overfitting and Double Descent Peter Bartlett (UC Berkeley)
10:15 am–10:30 am		Toward a Theory of Optimization for Over-Parameterized Systems of Non-Linear Equations: The Lessons of Deep Learning Misha Belkin (UC San Diego)
10:30 am–10:45 am		Overparameterization in Regression vs Classification Problems Anant Sahai (UC Berkeley)
10:45 am–11:00 am		Two Vignettes on Narrow Networks Matus Telgarsky (University of Illinois)
11:00 am–11:15 am		Finite vs. Infinite Neural Networks Jascha Sohl-Dickstein (Google Brain)
11:15 am–12:15 pm		Gather.town
12:15 pm–1:00 pm		Break
1:00 pm–1:15 pm		The Large Learning Rate Phase of Deep Learning Yasaman Bahri (Google Brain)
1:15 pm–1:30 pm		Finite-Sample Analysis of Interpolating Linear Classifiers in the Overparameterized Regime Phil Long (Google)
1:30 pm–1:45 pm		Chasing the Long Tail: What Neural Networks Memorize and Why Vitaly Feldman (Apple)
1:45 pm–2:00 pm		On Uniform Convergence and Interpolation Learning Nati Srebo (Toyota Technical Institute at Chicago)
2:00 pm–2:15 pm		Approximation Schemes for ReLU Regression Surbhi Goel (Microsoft Research)
2:15 pm–3:15 pm		Gather.town

Tuesday, August 11th, 2020
10:00 am–10:15 am		Reverse Isoperimetric Inequalities for Adversarial Machine Learning Varun Jog (University of Wisconsin, Madison)
10:15 am–10:30 am		Precise Tradeoffs in Adversarial Training Mahdi Soltanolkotabi (University of Southern California)
10:30 am–10:45 am		Extracting Robust and Accurate Features via a Robust Information Bottleneck Po-Ling Loh (University of Wisconsin, Madison)
10:45 am–11:00 am		Neural Perceptual Adversarial Robustness: Generalizable Defenses to Unforeseen Attacks Soheil Feizi (University of Maryland)
11:00 am–11:15 am		A Generalized Neural Tangent Kernel Analysis for Two-layer Neural Networks Quanquan Gu (UCLA)
11:15 am–12:15 pm		Gather.town
12:15 pm–1:00 pm		Break
1:00 pm–1:15 pm		The Implicit Regularization of Stochastic Gradient Flow for Least Squares Edgar Dobriban (University of Pennsylvania)
1:15 pm–1:30 pm		The Complexity of Non-Convex Stochastic Optimization Dylan Foster (MIT)
1:30 pm–1:45 pm		Initialization Scale vs Training Accuracy Effects on Implicit Bias in Deep Linear Classification Suriya Gunasekar (Microsoft Research)
1:45 pm–2:00 pm		Analytic Characterization of the Hessian in Shallow ReLU Models: A Tale of Symmetry Yossi Arjevani (New York University)
2:00 pm–2:15 pm		Beyond Linearization: On Quadratic and Higher-order Taylor Models of Wide Neural Networks Yu Bai (Salesforce Research)
2:15 pm–3:15 pm		Gather.town

Wednesday, August 12th, 2020
10:00 am–10:15 am		A Trichotomy of Rates in Supervised Learning Amir Yehudayoff (Technion)
10:15 am–10:30 am		Contrastive Estimation Reveals Topic Posterior Information to Linear Models Daniel Hsu (Columbia University)
10:30 am–10:45 am		Beyond Signal Propagation: Is Feature Diversity Necessary in Deep Neural Network Initialization? Daniel Soudry (Technion)
10:45 am–11:00 am		Few-Shot Learning via Learning the Representation, Provably Qi Lei (Princeton University)
11:00 am–11:15 am		How to Make Our Models Learn (the Right Things)? Aleksander Madry (MIT)
11:15 am–12:15 pm		Gather.town
12:15 pm–1:00 pm		Break
1:00 pm–1:15 pm		Generalization via Derandomization With an Application to Interpolating Predictors Gintare Karolina Dziugaite (Element AI)
1:15 pm–1:30 pm		How Does Our Mind Store Information in Memory? Rina Panigrahy (Google Research)
1:30 pm–1:45 pm		Fine-Tuning Language Models With Human Feedback Paul Christiano (OpenAI)
1:45 pm–2:00 pm		Equilibrium in Nonconvex-Nonconcave Min-Max Optimization and GANs Nisheeth Vishnoi (Yale)
2:00 pm–2:15 pm		Anatomy of Catastrophic Forgetting: Hidden Representations and Task Semantics Maithra Raghu (Cornell University)
2:15 pm–2:30 pm		Data Augmentation as Stochastic Optimization Boris Hanin (Princeton)
2:30 pm–3:15 pm		Gather.town

Thursday, August 13th, 2020
10:00 am–10:15 am		Tight Bounds on Minimax Regret under Logarithmic Loss via Self-Concordance Daniel Roy (University of Toronto)
10:15 am–10:30 am		The Intriguing Role of Module Criticality in the Generalization of Deep Networks Hanie Sedghi (Google Brain)
10:30 am–10:45 am		Predicting What You Already Know Helps: Provable Self-Supervised Learning Jason Lee (Princeton University)
10:45 am–11:00 am		Probabilistic Approximate Variants of Dimensional and Margin Complexity Pritish Kamath (Toyota Technological Institute at Chicago)
11:00 am–11:15 am		Shape Matters: Understanding the Implicit Bias of the Noise Covariance Tengyu Ma (Stanford University)
11:15 am–12:15 pm		Gather.town
12:15 pm–1:00 pm		Break
1:00 pm–2:15 pm		Group Discussion