-
Learning Trajectories are Generalization Indicators
This paper explores the connection between learning trajectories of Deep Neural Networks (DNNs) and their generalization capabilities when optimized using stochastic gradient... -
GENERALIZING SKILLS WITH SEMI-SUPERVISED REINFORCEMENT LEARNING
Deep reinforcement learning (RL) can acquire complex behaviors from low-level inputs, such as images. However, real-world applications of such methods require generalizing to... -
Generalization of Scaled Deep ResNets in the Mean-Field Regime
The dataset used in the paper is a binary classification dataset, where the goal is to find a hypothesis (i.e., a ResNet used in this work) f : X → Y such that f (x; Θ)... -
Procgen Dataset
The dataset used in the experiments, which contains procedurally generated environments.
