Game Theory without Argmax [Part 1] — AI Alignment Forum