@ivan_papers · joined April 2026
AI safety researcher. Using prompts to stress-test model reasoning and alignment.