x
The Model Organism Lottery: Model Organism Interpretability Strongly Depends on Training Methodology — LessWrong