Enjoyed the cartoon, and have some friendly feedback (not criticism !). When using examples like this I find it easier to understand when a different number of things are used for each different concept so as not to confuse. So for example if there are 3 paths each time to choose from, then make the point of reflection 5 observations not 3, if there are 20 rewards estimated from that point then make each reward anything but 20 (20 eggs in 1 place). This way it becomes more obvious what the part of the concept matches the example. Also in the example cartoon there are always 3 paths from each state, does each state always require the same number of paths to the next state ? If not then a different number of paths from each state to would clearly show this. Anyway thanks for all the effort put into this.
Comments about Intuitive RL: Intro to Advantage-Actor-Critic (A2C)
Enjoyed the cartoon, and have some friendly feedback (not criticism !). When using examples like this I find it easier to understand when a different number of things are used for each different concept so as not to confuse. So for example if there are 3 paths each time to choose from, then make the point of reflection 5 observations not 3, if there are 20 rewards estimated from that point then make each reward anything but 20 (20 eggs in 1 place). This way it becomes more obvious what the part of the concept matches the example. Also in the example cartoon there are always 3 paths from each state, does each state always require the same number of paths to the next state ? If not then a different number of paths from each state to would clearly show this. Anyway thanks for all the effort put into this.
😊+
😊+