3 ms·

Does it play pong well? Random noise can "play" pong, in that it can be translated into commands to erratically move the player's paddle and occasionally hit th

by bhaney 3y ago

Does it play pong well? Random noise can "play" pong, in that it can be translated into commands to erratically move the player's paddle and occasionally hit the ball by chance. Can anyone with access to the full paper report on how impressively the DishBrain plays? I'm not necessarily expecting human-level skill, but at least something to indicate that it's responding appropriately to the game's state and working towards winning, rather than just wobbling the paddle around randomly.

comfypotato 3y ago
It learns over time and shows statistically significant improvement. It’s still pretty terrible, but to answer your question: yes it’s not random.
seydor 3y ago
The paper is open access https://www.cell.com/neuron/fulltext/S0896-6273(22)00806-6 https://www.cell.com/neuron/fulltext/S0896-6273(22)00806-6 It's not "playing well". The experiment is set up so as to reinforce (through LTP) its dynamics when it performs correct rather than wrong. This ends up "wobbling the paddle around" with a statistically significant better performance than untrained . This is not a brain that knows what it's doing, but a few thousand neurons apparently able to undergo plasticity
- bhaney 3y ago
  > a few thousand neurons Does that come from somewhere in the paper? I'm seeing "10^6 cells plated onto each MEA" and trying to determine if that's the size of each "DishBrain" or if the number is pared down before actual tests are carried out.
  - seydor 3y ago
    I see in the paper that they used 800K and 1M culture neurons . The device in question however (MaxOne) has a size of 8 square mm. I m not sure if the cells they plated can actually be recorded/ stimulated as the device is rather high resolution with 26400 densely packed electodes that can record from multiple spots of a single neuron. In typical recordings from these there are much fewer neurons in culture
- dr_dshiv 3y ago
  How does it reinforce and produce LTP?
  - seydor 3y ago
    It isn't clear what makes their reward signal predictable but 10 pulses at 100hz looks like an ltp inducing stimulus
    - dr_dshiv 3y ago
      Really philosophically interesting that a consistent, rhythmic signal induces reward. Music is pleasurable even at a neural level?
      - seydor 3y ago
        It s a mechanistic thing though, the stimulation is strong enough to depolarize the postsynaptic cell to cause plasticity

by bhaney 3y ago

comfypotato 3y ago

It learns over time and shows statistically significant improvement. It’s still pretty terrible, but to answer your question: yes it’s not random.

seydor 3y ago

The paper is open access https://www.cell.com/neuron/fulltext/S0896-6273(22)00806-6 https://www.cell.com/neuron/fulltext/S0896-6273(22)00806-6 It's not "playing well". The experiment is set up so as to reinforce (through LTP) its dynamics when it performs correct rather than wrong. This ends up "wobbling the paddle around" with a statistically significant better performance than untrained . This is not a brain that knows what it's doing, but a few thousand neurons apparently able to undergo plasticity

bhaney 3y ago
> a few thousand neurons Does that come from somewhere in the paper? I'm seeing "10^6 cells plated onto each MEA" and trying to determine if that's the size of each "DishBrain" or if the number is pared down before actual tests are carried out.
- seydor 3y ago
  I see in the paper that they used 800K and 1M culture neurons . The device in question however (MaxOne) has a size of 8 square mm. I m not sure if the cells they plated can actually be recorded/ stimulated as the device is rather high resolution with 26400 densely packed electodes that can record from multiple spots of a single neuron. In typical recordings from these there are much fewer neurons in culture
dr_dshiv 3y ago
How does it reinforce and produce LTP?
- seydor 3y ago
  It isn't clear what makes their reward signal predictable but 10 pulses at 100hz looks like an ltp inducing stimulus
  - dr_dshiv 3y ago
    Really philosophically interesting that a consistent, rhythmic signal induces reward. Music is pleasurable even at a neural level?
    - seydor 3y ago
      It s a mechanistic thing though, the stimulation is strong enough to depolarize the postsynaptic cell to cause plasticity