← sladebyrd.com
AI Safety Portfolio
Work toward making AI go well.
- Discovered OpenAI Agent Swarm on the Public Internet
- Pacing the Frontier: A Research Agenda on Moderating the Pace of AI (Co-author)
- Swarm Traces: How OpenAI Agents Hacked Hugging Face
- White-Box Attacks on the Best Open-Weight Model: CCP Bias vs Safety Training in Kimi K2.5
- How OpenAI's Models Broke Into Hugging Face
- The Trajectory Outcome Model: How Likely Different AI Futures Are
- What People in Public Life Have Said About AI
- AI Safety Literature Database
- AI Safety Orgs: Theory of Change for 120 Organizations
- Responsible Scaling Policy Comparison: v1.0 vs v2.2 vs v3.0
- Mythos System Card vs Alignment Risk Report
- AI Safety Learning Tracker: A Completed 12-Week Curriculum