Toggle Poster Visibility
Spotlight
Wed Jul 20 11:30 PM -- 11:35 PM (KST) @ Ballroom 1 & 2
Towards understanding how momentum improves generalization in deep learning
Spotlight
Wed Jul 20 11:35 PM -- 11:40 PM (KST) @ Ballroom 1 & 2
What Can Linear Interpolation of Neural Network Loss Landscapes Tell Us?
Spotlight
Wed Jul 20 11:40 PM -- 11:45 PM (KST) @ Ballroom 1 & 2
Deep equilibrium networks are sensitive to initialization statistics
Spotlight
Wed Jul 20 11:45 PM -- 11:50 PM (KST) @ Ballroom 1 & 2
Scaling-up Diverse Orthogonal Convolutional Networks by a Paraunitary Framework
Spotlight
Wed Jul 20 11:50 PM -- 11:55 PM (KST) @ Ballroom 1 & 2
Stability Based Generalization Bounds for Exponential Family Langevin Dynamics
Spotlight
Wed Jul 20 11:55 PM -- 12:00 AM (KST) @ Ballroom 1 & 2
Local Augmentation for Graph Neural Networks
Spotlight
Thu Jul 21 12:00 AM -- 12:05 AM (KST) @ Ballroom 1 & 2
On Non-local Convergence Analysis of Deep Linear Networks
[
Paper PDF]
Oral
Thu Jul 21 12:05 AM -- 12:25 AM (KST) @ Ballroom 1 & 2
Adaptive Inertia: Disentangling the Effects of Adaptive Learning Rate and Momentum
Spotlight
Thu Jul 21 12:25 AM -- 12:30 AM (KST) @ Ballroom 1 & 2
Diversified Adversarial Attacks based on Conjugate Gradient Method
Spotlight
Thu Jul 21 12:30 AM -- 12:35 AM (KST) @ Ballroom 1 & 2
On the Optimization Landscape of Neural Collapse under MSE Loss: Global Optimality with Unconstrained Features
Spotlight
Thu Jul 21 12:35 AM -- 12:40 AM (KST) @ Ballroom 1 & 2
On the Equivalence Between Temporal and Static Equivariant Graph Representations
Spotlight
Thu Jul 21 12:40 AM -- 12:45 AM (KST) @ Ballroom 1 & 2
Robust Training under Label Noise by Over-parameterization
Spotlight
Thu Jul 21 12:45 AM -- 12:50 AM (KST) @ Ballroom 1 & 2
Implicit Bias of the Step Size in Linear Diagonal Neural Networks
[
Paper PDF]
Spotlight
Thu Jul 21 12:50 AM -- 12:55 AM (KST) @ Ballroom 1 & 2
Extended Unconstrained Features Model for Exploring Deep Neural Collapse
Spotlight
Thu Jul 21 12:55 AM -- 01:00 AM (KST) @ Ballroom 1 & 2
Score-Guided Intermediate Level Optimization: Fast Langevin Mixing for Inverse Problems
Successful Page Load