Strong Generalization

Strong generalization, the ability of machine learning models to perform well on unseen data, is a central objective in current research. Active areas of investigation include improving the robustness of self-supervised learning, understanding the optimization dynamics of transformers and other architectures (including CNNs and RNNs), and developing methods to enhance generalization through data augmentation, regularization techniques (e.g., logical regularization, consistency regularization), and improved training strategies (e.g., few-shot learning, meta-learning). These advancements are crucial for building reliable and adaptable AI systems across diverse applications, from image classification and natural language processing to healthcare and robotics.

Papers

November 15, 2023

November 12, 2023

November 11, 2023

November 10, 2023

Domain Generalization by Learning from Privileged Medical Imaging Information
Steven Korevaar, Ruwan Tennakoon, Ricky O'Brien, Dwarikanath Mahapatra, Alireza Bab-Hadiasha
Deep Learning Model Strong Generalization Medical Image Domain Generalization Privileged Information

November 9, 2023

November 8, 2023

Data Factors for Better Compositional Generalization
Xiang Zhou, Yichen Jiang, Mohit Bansal
Strong Generalization Generalization Performance Compositional Generalization Generalization Problem Characterization Data Data Complexity

November 6, 2023

SoK: Memorisation in machine learning
Dmitrii Usynin, Moritz Knolle, Georgios Kaissis
Machine Learning Strong Generalization Machine Learning Model Differential Privacy Model Generalization Memorisation Profile

November 5, 2023

November 3, 2023

Strong Generalization

Papers

Generalizable Imitation Learning Through Pre-Trained Representations

Improving Generalization of Drowsiness State Classification by Domain-Specific Normalization

Comparing Generalization in Learning with Limited Numbers of Exemplars: Transformer vs. RNN in Attractor Dynamics

On learning spatial sequences with the movement of attention

Federated Learning for Generalization, Robustness, Fairness: A Survey and Benchmark

Understanding Grokking Through A Robustness Viewpoint

Understanding Generalization via Set Theory

Minimum Description Length Hopfield Networks

Domain Generalization by Learning from Privileged Medical Imaging Information

The Paradox of Noise: An Empirical Study of Noise-Infusion Mechanisms to Improve Generalization, Stability, and Privacy in Federated Learning

Generalization in medical AI: a perspective on developing scalable models

Data Factors for Better Compositional Generalization

Selective Visual Representations Improve Convergence and Generalization for Embodied AI

Outliers with Opposing Signals Have an Outsized Effect on Neural Network Optimization

Principles from Clinical Research for NLP Model Generalization

SoK: Memorisation in machine learning

Robust Generalization Strategies for Morpheme Glossing in an Endangered Language Documentation Context

Learning Class and Domain Augmentations for Single-Source Open-Domain Generalization

RT-Trajectory: Robotic Task Generalization via Hindsight Trajectory Sketches

Tuning-less Object Naming with a Foundation Model