pith. sign in

Finding dori: Memorization in text-to-image diffusion mod- els is not local

2 Pith papers cite this work. Polarity classification is still indexing.

2 Pith papers citing it

fields

cs.CV 1 cs.LG 1

years

2026 2

verdicts

UNVERDICTED 2

representative citing papers

Memorization In Stable Diffusion Is Unexpectedly Driven by CLIP Embeddings

cs.CV · 2026-04-06 · unverdicted · novelty 6.0

Memorization in Stable Diffusion is driven by the structural duplication of the CLIP <eot> embedding inside <pad> tokens, which causes over-reliance on that vector; simple inference-time masking or token replacement suppresses it without quality loss.

citing papers explorer

Showing 2 of 2 citing papers.

  • Memorization In Stable Diffusion Is Unexpectedly Driven by CLIP Embeddings cs.CV · 2026-04-06 · unverdicted · none · ref 18 · internal anchor

    Memorization in Stable Diffusion is driven by the structural duplication of the CLIP <eot> embedding inside <pad> tokens, which causes over-reliance on that vector; simple inference-time masking or token replacement suppresses it without quality loss.

  • Diffusion Models Memorize in Training -- and Generalize in Inference cs.LG · 2026-03-12 · unverdicted · none · ref 41 · internal anchor

    Diffusion models overfit denoising loss at intermediate noise but generalize in inference as model error smooths the flow field and sampling paths avoid memorized noisy training data.