flâneur

Illuminating the Three Dogmas of RL under Evolutionary Light (Mani Hamidi) - Sensorimotor AI Journal Club

sensorimotorai.github.io · 3,785 words · saved by 1 readers

Last week, David Abel argued that RL’s central concepts (agent, learning, reward) need more precise definitions. He identified three dogmas limiting the field’s scientific ambitions. But he left one question conspicuously underdeveloped: what is adaptation, exactly?

Last week, David Abel argued that RL’s central concepts (agent, learning, reward) need more precise definitions. He identified three dogmas limiting the field’s scientific ambitions. But he left one question conspicuously underdeveloped: what is adaptation, exactly? This week, Mani Hamidi from the University of Tübingen picked up where Dave left off. His response paper offers evolutionary theory as a concrete paradigm that can address two of the three dogmas: it gives adaptation (Dogma 2) algorithmic substance through open-ended novelty search, and it complicates the reward hypothesis (Dogma 3

saved by

related reading