Backend/infrastructure engineer wandering into mechanistic interpretability and AI safety.
Interpretability, control, distributed systems, old poems, 4X games, experiments where nature gets to say “no,” and walls that don’t actually stop anyone. GitHub