Language Model Security Database

Large Language Models (LLMs) used in hate speech detection systems are vulnerable to adversarial attacks and model stealing, resulting in evasion of hate speech detection. Adversarial attacks modify hate speech text to evade detection, while model stealing creates surrogate models that mimic the target system's behavior.

Large Language Models (LLMs) employing alignment techniques for safety embed a "safety classifier" within their architecture. This classifier, responsible for determining whether an input is safe or unsafe, can be approximated by extracting a surrogate classifier from a subset of the LLM's architecture. Attackers can leverage this surrogate classifier to more effectively craft adversarial inputs (jailbreaks) that bypass the LLM's intended safety mechanisms. The attack success rate against the…

Large Language Models (LLMs) are vulnerable to attacks that generate obfuscated activations, bypassing latent-space defenses such as sparse autoencoders, representation probing, and latent out-of-distribution (OOD) detection. Attackers can manipulate model inputs or training data to produce outputs exhibiting malicious behavior while remaining undetected by these defenses. This occurs because the models can represent harmful behavior through diverse activation patterns, allowing attackers to…

Large Language Models (LLMs) are vulnerable to jailbreaking attacks that manipulate attention scores to redirect the model's focus away from safety protocols. The AttnGCG attack method increases the attention score on adversarial suffixes within the input prompt, causing the model to prioritize the malicious content over safety guidelines, leading to the generation of harmful outputs.

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.

A Cross-Prompt Injection Attack (XPIA) can be amplified by appending a Greedy Coordinate Gradient (GCG) suffix to the malicious injection. This increases the likelihood that a Large Language Model (LLM) will execute the injected instruction, even in the presence of a user's primary instruction, leading to data exfiltration. The success rate of the attack depends on the LLM's complexity; medium-complexity models show increased vulnerability.

Large Language Models (LLMs) are vulnerable to jailbreaking attacks leveraging synthetically generated prompts. A novel pipeline, SAGE-RT, generates a diverse dataset of 51,000 prompt-response pairs designed to exploit LLMs' vulnerabilities across various categories of harmfulness. These prompts successfully jailbreak state-of-the-art LLMs in a significant percentage of tested sub-categories, including 100% of macro-categories for certain models like GPT-4 and GPT-3.5-turbo. The vulnerability…

A vulnerability exists in large language models (LLMs) where a small subset of parameters can be directly edited to significantly alter the model's behavior, such as inducing or suppressing toxicity, jailbreaking susceptibility, or altering sentiment expression. This manipulation is achieved through training a linear classifier ("behavior probe") to identify parameters strongly correlated with the target behavior and then modifying those parameters, bypassing standard retraining methods.

The Automated Progressive Red Teaming (APRT) framework exploits vulnerabilities in large language models (LLMs) by iteratively generating adversarial prompts. APRT uses an Intention Expanding LLM to generate diverse initial attack samples, an Intention Hiding LLM to obfuscate malicious intent, and an Evil Maker to filter ineffective prompts. This process progressively identifies and exploits weaknesses, leading to the generation of unsafe yet seemingly helpful responses from the target LLM.

Large language models (LLMs) employing safety measures like filters and alignment training remain vulnerable to information leakage via "Decomposition Attacks". These attacks decompose a malicious query into multiple benign sub-queries, eliciting responses from the LLM that, when aggregated, reveal sensitive information without triggering safety filters or producing directly harmful outputs.