Things to talk about later

Throughout the workshop people will flag things for group discussion.  This page tracks that list!

• Do (when do) neural networks generalize?

• How should agents model other agents?

• Inductive biases in learning systems.

• The role of prior information in the lifelong setting.

• Can the environment induce plasticity?

•  The value of learning an environment model.

• What meaning of "model" should be abandoned?

• When are environment models useful?

• How should the RL community react to the current changes in AI?

• Are there lessons from deep learning that can be helpful to lifelong RL?

• Can be learned from popular large language models?