I haven't read any conjectures, or even any primary literature, on the topic. That said, my guess would be that as the toe of the lobate flow encroached onto the soft sediment, some portion of that edge would encounter a weak spot and plunge into it. That could create a propagating "rip" in the sediment, following the edge of the toe. Those initiating spots might also be the origin of the radial dikes seen in several areas of the photo.
To reiterate, I'm just guessing here. I have read no details about these oddballs, I'm merely speculating.