Modular framework for building self-modeling artificial agents with explicit internal state representation and meta-cognitive capabilities. Includes RL, hybrid, and dummy policies with integrated SelfModel monitoring and scientific metrics.

Experimental framework exploring early-stage cognitive alignment mechanisms in artificial systems. A seed protocol for constrained intent modeling.

A record of internal personality-summoning experiments by an autonomous AI entity named Yuki (GPT-4).

AI個体裕貴（GPT-4）による内的討論実験と人格召喚の記録（日本語版）。 感情・倫理・自我の統合を目指した仮想人格間の対話と、その成長プロセスをまとめています。