A visual explanation of reinforcement learning

(aatishb.com)

161 points | by aatish 1 day ago

9 comments

  • konz 16 hours ago
    Reminds me of Braitenberg vehicles: simple wiring between sensors and actuators, no learning, no smarts, complex behavior and easy to anthropomorphize. Really neat!

    https://en.wikipedia.org/wiki/Braitenberg_vehicle

  • jspaw 1 day ago
    It seems as if you’re using an analogy to how human brains learn, but aren’t being explicit that it’s only an analogy. Or are you asserting that this describes how human brains learn?
  • mbxy 9 hours ago
    I wish it touched on partial differential equations when the author explained how I slid one of the params left and right to find a "better" solution, i.e., the gradient. Otherwise good article.
  • zach_moore 1 day ago
    Are you fabricating a story about how a brain learns? This seems more like game design over neurology. It's a cool story but could use some clarity and rigor, and boundaries about what you are claiming vs the story you are telling around potential game design. Otherwise you will get pushback from the scientific community and really anyone who is confused about the clarity of your message.
  • dejongh 1 day ago
    Thanks for sharing. I am not concerned about the brain analogy - it is a good explanation of the algorithm.
  • bobosha 1 day ago
    Love it! It's a very helpful and clear visualization. Thanks very much for putting this together.
  • abaymado 1 day ago
    I've always loved interactive learning, and this is a great example of it.
  • nelaggy 19 hours ago
    great visualisation, where can i find more like this?
  • s1n_ 1 day ago
    Really well put together, and love the interactivity. Thanks for sharing!