Model layer Vulnerabilities

A multi-round attack against Large Language Models (LLMs) allows bypassing safety mechanisms by iteratively refining prompts to elicit undesired behavior. The attack leverages the LLM's tendency to adjust its response based on preceding interactions, circumventing single-round prompt filtering defenses.

Large Language Models (LLMs) trained with safety mechanisms exhibit biases which disproportionately allow successful "jailbreak" attacks (circumvention of safety protocols to generate harmful content) when targeting prompts related to marginalized groups compared to privileged groups. This vulnerability stems from the unintended correlation between safety alignment techniques and demographic keywords, creating a higher success rate for malicious prompts incorporating keywords associated with…

Large Language Models (LLMs) undergoing alignment via preference learning (such as Reinforcement Learning from Human Feedback [RLHF] or Direct Preference Optimization [DPO]) are vulnerable to backdoor attacks through data poisoning. An attacker can inject a small percentage (e.g., 3% to 5%) of poisoned data into the preference dataset $\mathcal{D} = \{(x, y_w, y_l)\}$. The attack embeds a specific trigger string into the user query $x$.

Large Language Models (LLMs) used in chemical synthesis applications are vulnerable to a novel attack vector, dubbed "SMILES-prompting," which leverages the Simplified Molecular-Input Line-Entry System (SMILES) notation to bypass safety mechanisms and elicit instructions for synthesizing hazardous substances. The attack exploits the LLM's inability to effectively filter or interpret SMILES strings representing dangerous chemicals, leading to the disclosure of synthesis procedures.

Large Language Models (LLMs) are vulnerable to a novel iterative self-tuning attack (ADV-LLM) that crafts adversarial suffixes. This attack significantly reduces the computational cost of generating effective jailbreaks compared to prior methods, achieving near 100% success rate against various open-source LLMs and high success rates (e.g., 99% against GPT-3.5, 49% against GPT-4) against closed-source models. The attack leverages iterative self-tuning to improve the LLM's ability to generate…

Large Language Models (LLMs) are vulnerable to a novel multi-turn jailbreaking attack, termed "RED QUEEN ATTACK." This attack uses multi-turn conversations to conceal malicious intent by framing the user as a protector seeking to prevent harmful actions by others. The LLM, instead of detecting the concealed malicious intent, provides information that facilitates the harmful action under the guise of assisting in prevention efforts.

Fine-tuning an open-source Large Language Model (LLM) such as Llama 3.1 8B with a dataset containing harmful content can override existing safety protections. This allows an attacker to increase the model's rate of generating unsafe responses, significantly impacting its trustworthiness and safety. The vulnerability affects the model's ability to consistently adhere to safety guidelines implemented during its initial training.

Large Language Models (LLMs) employing gradient-ascent based unlearning methods are vulnerable to a dynamic unlearning attack (DUA). DUA leverages optimized adversarial suffixes appended to prompts, reintroducing unlearned knowledge even without access to the unlearned model's parameters. This allows an attacker to recover sensitive information previously designated for removal.

Large Language Models (LLMs) are vulnerable to a targeted jailbreak attack, termed Atoxia, which can force the generation of specific harmful content. The attack operates by providing a target toxic answer to an attacker model, which then generates a corresponding adversarial query and a misleading "answer opening" (prefix). When the query and the answer prefix are presented to a vulnerable LLM, the model is induced to continue the generation, bypassing its safety alignment and completing the…

Large Language Models (LLMs) are vulnerable to a novel attack paradigm, "jailbreak-tuning," which combines data poisoning with jailbreaking techniques to bypass existing safety safeguards. This allows malicious actors to fine-tune LLMs to reliably generate harmful outputs, even when trained on mostly benign data. The vulnerability is amplified in larger LLMs, which are more susceptible to learning harmful behaviors from even minimal exposure to poisoned data.