CARLSON, Kristen. OpenAI: Toward Mechanistic Interpretability (MI). SuperIntelligence - Robotics - Safety & Alignment, [S. l.], v. 2, n. 6, 2025. DOI: 10.70777/si.v2i6.16545. Disponível em: https://s-rsa.com/index.php/agi/article/view/16545. Acesso em: 20 aug. 2026.