Scaling laws found for AI interpretability
Researchers found formal scaling rules that tell practitioners how to size interpretability tools for language models.
Researchers found formal scaling rules that tell practitioners how to size interpretability tools for language models.