SAD modifies the denoising process in text diffusion models to enforce safety constraints at inference time, reducing unsafe generations while preserving quality and diversity.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
DR-Smoothing introduces a disrupt-then-rectify prompt processing scheme into smoothing defenses, delivering tight theoretical bounds on success probability against both token- and prompt-level jailbreaks.
citing papers explorer
-
The Safety-Aware Denoiser for Text Diffusion Models
SAD modifies the denoising process in text diffusion models to enforce safety constraints at inference time, reducing unsafe generations while preserving quality and diversity.
-
Guaranteed Jailbreaking Defense via Disrupt-and-Rectify Smoothing
DR-Smoothing introduces a disrupt-then-rectify prompt processing scheme into smoothing defenses, delivering tight theoretical bounds on success probability against both token- and prompt-level jailbreaks.