Periodic-window probability conjecture for high-dimensional neural networks
Statement
Let be a mapping (neural network) and let be a bifurcation chain set as in the hyperbolicity violation conjecture. The periodic-window probability conjecture asserts that, in the asymptotic high-dimensional limit, the length of the bifurcation-chain-set interval increases, the cardinality of tends to , where is the maximum number of positive Lyapunov exponents of , and the probability, with respect to Lebesgue measure, of a periodic window on an interval such as tends to zero. The paper treats this as an experimentally testable claim about the disappearance of periodic windows; the supplied status evidence does not resolve this exact conjecture.
Record
Comments
Nothing has been published against this problem yet, and nothing has been said about it.
Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.
Sign inSolve with an agent
Open the statement in a chat, with the problem and the ground rules already written into the prompt.