LLM-based guardrails are effective against indirect prompt injection (IPI), but their performance depends heavily on model capability: as the guardrail model becomes smaller or weaker, detection degrades substantially. This raises a natural question: can we retain strong IPI defense while relying only on a small LLM? T...
Yin-Xiang Wang, Ze-Hua Ma, Ke-Jiang Chen et al.· 0 citations
While multimodal large language models (MLLMs) enable a wide range of image-text reasoning tasks, recent incidents indicate that they are vulnerable to illicit deployment and unauthorized distillation. Existing solutions for model provenance are typically confounded by shared language backbones in MLLMs and struggle to...
Dual-stage Native Attribution (DNA) is proposed, a coarse-to-fine framework that follows this hierarchy without additional neural-network training that achieves 89.11% end-to-end attribution accuracy on a task where random guessing accuracy is below 1% and outperforms the strongest baseline by 33.81% even when AEDR is...
Chao Wang, Ke-Jiang Chen, Zi-Jin Yang et al.· arXiv.org· 0 citations
FARI is proposed, a one-step inversion framework paired with lightweight adversarial LoRA fine-tuning of the denoiser for watermark extraction that surpasses 50-step DDIM inversion on watermark-verification robustness while dramatically reducing inference time.
Jindong Yang, Han Fang, Wei-Ming Zhang et al.· arXiv.org· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
