“…In contrast, in the fast diffusion case m ∈ (0, 1), the path-by-path wellposedness of (1.1) and the related problem of the existence of a corresponding stochastic flow and random dynamical system have proven to be notoriously difficult and have been posed as open questions in several works, e.g. Barbu and Röckner [4,5,6] and the second author [11,13]. In addition, even in the porous medium case m ∈ (1, ∞), for general initial data u 0 ∈ L 2 (U ), path-by-path solutions could only be obtained in a limiting sense in [4,6,11], thus lacking a characterization in terms of (generalized) solutions to (1.1).…”