Another visualization of the #Collatz sequence. This time highlighting some streak-like structures. Btw, the cellular automaton for the Collatz sequence is just \[ s_{i,j}=(s_{i-1,j}\mathbin{\text{mod}} 2) - (s_{i,j+1}\mathbin{\text{mod}} 3)\] None of that seems to help for the Collatz conjecture, but at least one can easily show that in any long cycles at some point about half of the leading binary digits must be `11000...`. #math