Can you post some good or interesting genomes?

[danbri]

I've run it for 1500+ generations and don't get beyond 3 steps. It would be interesting to see what the best strategies found have been. People could just post https://gist.github.com/ links in this issue thread...

[leshchenko1979]

I think this inability of the network to achieve desired results may be due to the following factors:

• High motorNoise impedes moving positive mutations to the next generations. I've tried lowering it to zero and achieved 3 steps and 12K score in ~40 generations. But couldn't get any further from there.
• The current speed of the human should be an input node of the network - as different moves are required to start going from standing still compared to the moves needed to continue walking

[Inthenew]

i got to 4 steps in like 50 gens