Chijioke Ugwuanyi

Steer­ing Black­mail Through a Model’s “Emo­tional State”

Schem­ing Evals Mislead in Both Directions

From 8B to Fron­tier: How Sys­tem Prompts Con­trol Whether AI Agents Black­mail, Leak, and Kill

Black­mail at 8 Billion Pa­ram­e­ters: Agen­tic Misal­ign­ment in Sub-Fron­tier Models

Repli­ca­tion of Koorndijk (2025): Differ­en­tial Com­pli­ance May Reflect Prompt Sen­si­tivity Rather Than Strate­gic Reasoning