SALLIE: Safeguarding Against Latent Language & Image Exploits

Computer Science > Cryptography and Security [Submitted on 6 Apr 2026] SALLIE: Safeguarding Against Latent Language & Image Exploits Guy Azov, Ofer Rivlin, Guy Shtar Large Language Models (LLMs) and Vision-Language Models (VLMs) remain highly vulnerable to textual and visual jailbreaks, as well as prompt injections (arXiv:2307.15043, Greshake et al., 2023, arXiv:2306.13213). Existing defenses often degrade performance through complex input transformations or treat multimodal threats as isolated problems (arXiv:2309.00614, arXiv:2310.03684, Zhang et al., 2025). To address the critical gap for a unified, modal-agnostic defense that mitigates both textual and visual threats simultaneously without degrading performance or requiring architectural modifications, we introduce SALLIE (Safeguarding Against Latent Language & Image Exploits), a lightweight runtime detection framework rooted in mechanistic interpretability (Lindsey et al., 2025, Ameisen et al., 2025). By integrating seamlessly into standard token-level fusion pipelines (arXiv:2306.13549), SALLIE extracts robust signals directly from the model's internal activations. At inference, SALLIE defends via a three-stage architecture: (1) extracting internal residual stream activations, (2) calculating layer-wise maliciousness scores using a K-Nearest Neighbors (k-NN) classifier, and (3) aggregating these predictions via a layer ensemble module. We evaluate SALLIE on compact, open-source architectures - Phi-3.5-vision-instruct (arXiv:2404.14219), SmolVLM2-2.2B-Instruct (arXiv:2504.05299), and gemma-3-4b-it (arXiv:2503.19786) - prioritized for practical inference times and real-world deployment costs. Our comprehensive evaluation pipeline spans over ten datasets and more than five strong baseline methods from the literature, and SALLIE consistently outperforms these baselines across a wide range of experimental settings. Comments: 18 pages, 4 figures, 7 tables. Preprint under review Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) ACM classes: I.2.7; I.2.10; K.6.5; I.2.6 Cite as: arXiv:2604.06247 [cs.CR]   (or arXiv:2604.06247v1 [cs.CR] for this version)   https://doi.org/10.48550/arXiv.2604.06247 Focus to learn more Submission history From: Ofer Rivlin [view email] [v1] Mon, 6 Apr 2026 16:29:05 UTC (566 KB) Access Paper: HTML (experimental) view license Current browse context: cs.CR < prev   |   next > new | recent | 2026-04 Change to browse by: cs cs.AI References & Citations NASA ADS Google Scholar Semantic Scholar Export BibTeX Citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Demos Related Papers About arXivLabs Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)