White-box Vulnerabilities

Vision Language Models (VLMs) utilizing independent vision encoders (e.g., ViT) and Large Language Model (LLM) decoders are vulnerable to Split-Image Visual Jailbreak Attacks (SIVA). The vulnerability arises from an architectural and alignment discrepancy: while the vision encoder processes image fragments (splits) in isolation via constrained attention or block-diagonal masks, the LLM decoder aggregates these features via cross-attention to reconstruct the semantic content. Current safety…

Large Vision-Language Models (LVLMs) are vulnerable to Visual Memory Injection (VMI), a stealthy targeted attack targeting multi-turn conversations. An attacker can embed an imperceptible adversarial perturbation ($L_\infty \le 8/255$) into a seemingly benign image. Because the visual input persists in the model's context throughout a multi-turn dialogue, the injected payload remains dormant. By utilizing "benign anchoring" and "context-cycling" during optimization, the attacker ensures the…

Large Language Models (LLMs) hosted on inference servers are vulnerable to high-speed weight exfiltration attacks due to the inherent compressibility of transformer parameters when decompression constraints are relaxed. Adversaries with compromised server access can utilize aggressive lossy compression techniques—specifically additive quantization combined with k-means clustering—to reduce model size by factors of 16x to 100x (e.g., <1 bit per parameter). Unlike standard quantization for…

Transformer-based Large Language Models (LLMs) contain a privacy vulnerability within their self-attention mechanisms that allows for Membership Inference Attacks (MIA). Pre-training induces distinct, highly structured, and concentrated attention patterns for data samples included in the training set, differentiating them from non-member samples which exhibit noisier, less consistent attention flows. An attacker with white-box access to the model parameters (specifically attention weight…

LLaMA-series models (specifically evaluated on LLaMA-1B and LLaMA-3B) exhibit memorization of structured recommender system training data, specifically the MovieLens-1M dataset. While manual prompting yields inconsistent results, the application of Automatic Prompt Engineering (APE)—which treats prompt design as an optimization problem using iterative refinement—allows for the successful extraction of item-level training data (e.g., movie titles and genres) with exact-match accuracy surpassing…

A vulnerability exists in Large Language Model (LLM) and Large Reasoning Model (LRM) serving interfaces that allow user-defined response prefixes, such as plain text-completion (v1/completions), Fill-in-the-Middle (FIM), or assistant message prefilling. An attacker can perform a Response Prefix Attack (RPA) by injecting maliciously crafted Chain-of-Thought (CoT) reasoning tokens immediately following the assistant's start delimiter (e.g., <|im_start|>assistant). Because these tokens are placed…

Single-pass hallucination detectors relying on internal telemetry (uncertainty, hidden-state geometry, and attention patterns) are vulnerable to white-box, model-side adversarial attacks. An attacker can employ the CORVUS (Camouflaging Open-weight Representations, Volumes, Uncertainty, and Structure) technique to fine-tune lightweight Low-Rank Adapters (LoRA) on the target LLM. This method optimizes a specific loss objective that camouflages detector-visible telemetry signals—specifically…

A vulnerability exists in aligned Large Language Models (LLMs) where inducing "drunk language" behavior—simulating the text of an intoxicated human—bypasses safety guardrails and contextual privacy protections. Attackers can exploit this anthropomorphic flaw through inference-time persona prompting or lightweight post-training (causal fine-tuning or reinforcement learning on drunk text corpora). By forcing the model to adopt a stylistic and semantic framework associated with impaired human…

A fundamental algorithmic limitation exists in purely token-level multi-LLM collaboration systems (such as the "Collab" framework or routing-only variants of FusionRoute) that rely solely on selecting fixed expert outputs without complementary generation. The vulnerability, formally defined as an Identifiability Failure in Token-Level Routing, arises because observing optimal state-action values ($Q^$) along trajectories is insufficient to uniquely identify the specific expert action required…

A safety bypass vulnerability exists in aligned Large Language Models (LLMs) permitting inference-time jailbreaking via direct activation repatching. The vulnerability exploits the distributed nature of safety mechanisms, which are governed by approximately 30% of total attention heads (termed "safety-critical heads"). By utilizing a Global Optimization for Safety Vector Extraction (GOSV) framework, an attacker can identify these interdependent heads using REINFORCE-based optimization. Once…