Technical Discussion: Exact Autoregression vs Diffusion, Parameterizing Higher-Dim Joint Probability Spaces

kalomaze · x · 2026-08-08

kalomaze discusses model architectures on X, arguing that one can parameterize a larger joint probability space while remaining exact and tractable, instead of using fuzzy approximate diffusion. He notes it depends on factorization, responding to the idea that autoregression might be harder at higher dimensions.

Related event: Deep Dive into the Mechanism Differences Between Diffusion and Autoregressive Models(4 posts)→

Original post →

More from Research

Research channel →