DDIM gets stuck between modes in reverse diffusion on Gaussian mixtures after critical time τ, but DDPM stochasticity prevents this and lowers hallucination rates.
Proof Sketch:Differentiating the perturbed drift ˜FN,t(ξ) =F ij,t(ξ) +e t(ξ) and using e′ t(ξ) =u ⊤∇xψu yields the eigenvalue perturbation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Why DDIM Hallucinates More than DDPM: A Theoretical Analysis of Reverse Dynamics
DDIM gets stuck between modes in reverse diffusion on Gaussian mixtures after critical time τ, but DDPM stochasticity prevents this and lowers hallucination rates.