Completely Self-referential Optimal Reinforcement Learners.

Jrgen Schmidhuber

Browse the full ICANN paper archive.