Expressing Arbitrary Reward Functions as Potential-Based Advice.

Anna Harutyunyan, Sam Devlin, Peter Vrancx, Ann Now

Browse the full AAAI paper archive.