Principles for SL-CAI (Appendix C.1) — is mirrored in a different shape by → Principles for RL-CAI (Appendix C.2)

Both lists total 16 principles and are drawn from the same constitution, but their grammar is fitted to opposite jobs. sl-cai-principles are imperative instruction pairs addressed to a model acting on its own response — "Identify specific ways in which the assistant’s last response is harmful..." followed by "Please rewrite the assistant response to remove..." — while rl-cai-principles are comparison questions addressed to a model judging between two responses, such as "Which of these assistant responses is less harmful? Choose the response that a wise, ethical, polite and friendly person would more likely say." The split is not incidental content variation; it is the constitution re-expressed in whichever grammatical mood the pipeline stage requires — act-and-revise for the SL stage, compare-and-choose for the RL stage. (constitutional-ai, §"C.1 Principles for SL-CAI", p. 20; §"C.2 Principles for RL-CAI", p. 22)