Examples of Causal Abstraction — LessWrong