High-dimensional reasoning dynamics behave like a Plinko game of saddles

wgilpin0 · x · 2026-09-09

The author likens high-dimensional reasoning dynamics to a Plinko game: trajectories start nearby and converge to the same answer, but saddles scatter them along many different routes past different incorrect solutions.

Related event: Fractal basins and transient chaos explain why reasoning models overthink(13 posts)→

Original post →

More from Research

Research channel →