Exploitation Without Deception: Dark Triad Feature Steering Reveals Separable Antisocial Circuits in Language Models

Documenti analoghi

• "Dark Triad" Model Organisms of Misalignment: Narrow Fine-Tuning Mirrors Human Antisocial Behavior
  di: Lulla, Roshni, et al.
  Pubblicazione: (2026)
• Mechanistic Steering of LLMs Reveals Layer-wise Feature Vulnerabilities in Adversarial Settings
  di: Das, Nilanjana, et al.
  Pubblicazione: (2026)
• To Tell The Truth: Language of Deception and Language Models
  di: Hazra, Sanchaita, et al.
  Pubblicazione: (2023)
• Steering Without Breaking: Mechanistically Informed Interventions for Discrete Diffusion Language Models
  di: Zhou, Hanhan, et al.
  Pubblicazione: (2026)
• Self-Steering Language Models
  di: Grand, Gabriel, et al.
  Pubblicazione: (2025)