Step-wise detection via a contrastive safety direction followed by remasking and adaptive steering reduces jailbreak success rates in diffusion language models to 0.64% while preserving output quality.
Guardnet: Graph-attention filtering for jailbreak defense in large language models.CoRR, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Adaptive Steering and Remasking for Safe Generation in Diffusion Language Models
Step-wise detection via a contrastive safety direction followed by remasking and adaptive steering reduces jailbreak success rates in diffusion language models to 0.64% while preserving output quality.