Introduces AIR, an asymmetric regularization that anchors open-ended safety prompts to verifiable ones via stop-gradient, improving invariance and accuracy when combined with group preference optimization.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.CL 3years
2026 3verdicts
UNVERDICTED 3representative citing papers
MultiBreak is a large diverse multi-turn jailbreak benchmark that achieves substantially higher attack success rates on LLMs than prior datasets and reveals topic-specific vulnerabilities in multi-turn settings.
UAG is a universal avoidance generation method that increases multi-branch diversity in diffusion and transformer models by penalizing output similarity, delivering up to 1.9x higher diversity with 4.4x speed and 1/64th the FLOPs of prior methods.
citing papers explorer
-
Towards Context-Invariant Safety Alignment for Large Language Models
Introduces AIR, an asymmetric regularization that anchors open-ended safety prompts to verifiable ones via stop-gradient, improving invariance and accuracy when combined with group preference optimization.
-
MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Evaluating LLM Safety
MultiBreak is a large diverse multi-turn jailbreak benchmark that achieves substantially higher attack success rates on LLMs than prior datasets and reveals topic-specific vulnerabilities in multi-turn settings.
-
A Universal Avoidance Method for Diverse Multi-branch Generation
UAG is a universal avoidance generation method that increases multi-branch diversity in diffusion and transformer models by penalizing output similarity, delivering up to 1.9x higher diversity with 4.4x speed and 1/64th the FLOPs of prior methods.