Readings

Readings complement lecture content and links will be periodically posted below.

corresponding lecture	readings
lecture 1, 8	Overcoming Catastrophic Forgetting in Neural Networks
lecture 1, 5	Continual Backprop: Stochastic Gradient Descent with Persistent Randomness
lecture 2	Adapting Bias by Gradient Descent: An Incremental Version of Delta-Bar-Delta
lecture 2	Gain Adaptation Beats Least Squares?
lecture 3, 4	Simple Agent, Complex Environment: Efficient Reinforcement Learning with Agent States
lecture 6	Toward a Formal Framework for Continual Learning
lecture 6	Mark Ring's Slides: My Take on Continual Learning
lecture 10	A Tutorial on Thompson Sampling
lecture 11	Efficient Continual Learning with Modular Networks and Task-Driven Priors
lecture 11	NEVIS’22: A Stream of 100 Tasks Sampled from 30 Years of Computer Vision Research
lecture 12, 13	Non-Stationary Bandit Learning via Predictive Sampling
lecture 13, 15	An Information-Theoretic Framework for Supervised Learning
lecture 17	Deciding What to Learn: A Rate-Distortion Approach

Supplementary background materials

supplementary materials

Supplementary Math Background

Reinforcement Learning: An Introduction

General surveys and papers we may cover later

surveys and papers

Embracing Change: Continual Learning in Deep Neural Networks

Towards Continual Reinforcement Learning: A Review and Perspectives

Efficient Continual Learning with Modular Networks and Task-Driven Priors

An Information-Theoretic Analysis of Compute-Optimal Neural Scaling Laws