The integration of external tools and untrusted data has transformed large language models (LLMs) into autonomous agents, but also exposes them to indirect prompt injection (IPI) attacks, where adversarial instructions in external content can manipulate tool use and compromise system behavior. Existing defenses use i...
Yuewantong Song, Guan-Hang Shi, Xue-Mei Luo et al.· Journal of computing and sec...· 0 citations
Internal safety scores judge a prompt before any text is generated, and they are validated by how well they separate harmful prompts from benign ones, so Active Attention Probing is introduced, which supplies a fixed content independent measurement coordinate.
Ming-Yu Luo, Ming Deng, Zilang Qiu et al.· 0 citations
These findings characterize the potential of single-engine diversity to detect wrong-result failures and select correct results, together with the execution costs and reference-validity constraints that shape its practical use.
Ze-Lin Wang, Ping Chen, Jin Wei et al.· Security and Safety· 0 citations
ActProbe, an internal-state-based framework for detecting and localizing poisoned segments in multi-source LLM inputs, is proposed and remains effective against defense-aware adaptive attacks and can protect black-box APIs through surrogate-based poisoned-segment removal.
Xue Tan, Chang-Hui Wang, Sanrui Yang et al.· 0 citations
ToSS (Token Oriented Repartitioning and Strategic Selection), a reliable authentication method using adaptive dual watermarking that divides vocabulary tokens into black and white sublists, enabling precise bit-level embedding of traceability information.
Zhongli Fang, Yi-Ran Chen, Ling-Yun Zhang et al.· 0 citations
AcMAS is proposed, an activation-based framework for malicious-behavior detection in MAS that significantly outperforms graph-based baselines against stealthy attacks, with generalization across diverse open-source LLM backbones, attack intensity, and MAS scale.
Haowen Xu, Xue Tan, Lei Ma et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.