deep reinforcement learning
1 statements across 1 episodes · 0 bullish · 0 bearish · 1 people on the record · first statement Aug 11, 2024 by Brian Dubois · across every show →
Everything said about deep reinforcement learning, oldest first
Aug 11, 2024 neutral
Dubois: Industrial reinforcement learning cannot generalize across factory tasks
“And that's really interesting because it learns to build long-term strategy. It can make decisions and things like that, but it's still so narrow. Like it's really capable, but only on the task that it's been trained on. Like if you even just put it on a diffe…”