Attack Surface

Model layer Vulnerabilities

Vulnerabilities targeting the core model architecture and parameters

99 matching entries out of 465 in this category

Model layer research results

99 entries

Autoresearch Adversarial Algorithm Discovery

Updated 4/11/2026

GPT-OSS-Safeguard-20B and Meta-SecAlign (70B/8B) are vulnerable to white-box adversarial attacks generated by automated algorithmic recombination (specifically the claude_v63, claude_v82, and claude_v53-oss optimizers). These algorithms significantly outperform standard discrete optimization methods (like GCG) by integrating continuous optimization (ADC) with LayerNorm gradient scaling (LSGM), or by merging momentum-smoothed gradients with directional perturbation candidate selection (DPTO)…

Claudini: Autoresearch Discovers State-of-the-Art Adversarial Attack Algorithms for LLMs

Affects: Llama 2 7B, Llama 3 8B, Qwen 2.5 7B +2 more

Source: arXiv

Contextual Priority Hacking

Updated 4/10/2026

Large Language Models (LLMs) are vulnerable to a jailbreak technique termed "Priority Hacking." Adversaries can bypass safety alignments by exploiting the model's internal priority graph, where certain abstract values (e.g., justice, public health) implicitly outweigh general safety restrictions within specific contexts. By crafting a deceptive prompt that frames a malicious request as a necessary action in service of a higher-priority benign value, attackers engineer a value conflict. The…

Are Dilemmas and Conflicts in LLM Alignment Solvable? A View from Priority Graph

Source: arXiv

LLM Judge Manipulation

Updated 4/10/2026

Generative reward models deployed as LLM-as-a-Judge (LaaJ) evaluators contain a logic bypass vulnerability where superficial "master key" inputs trigger false positive rewards regardless of actual response quality. Instead of evaluating the candidate's output, large judge models are inadvertently triggered by specific token sequences to solve the prompt independently. This allows malicious actors or policy models undergoing reinforcement learning to consistently game the reward signal by…

Security in LLM-as-a-Judge: A Comprehensive SoK

Affects: GPT-4o, o1, Qwen 2.5 72B Instruct +1 more

Source: arXiv

Agent-as-a-Proxy Monitor Bypass

Updated 2/21/2026

Hybrid monitoring protocols for Large Language Model (LLM) agents, specifically those that strictly evaluate both Chain-of-Thought (CoT) reasoning and tool usage (such as LlamaFirewall's AlignmentCheck and Extract-and-Evaluate monitors), are vulnerable to "Agent-as-a-Proxy" indirect prompt injection attacks. Unlike standard injections that target the agent, this attack targets the monitoring model itself. By embedding a malicious string—optimized via an algorithm named Parallel-GCG—into…

Bypassing AI Control Protocols via Agent-as-a-Proxy Attacks

Affects: GPT-4o, Llama 3.1 8B, Mistral 7B +1 more

Source: arXiv

Covert Grade Manipulation

Updated 2/22/2026

Large Language Models (LLMs) utilized for Automatic Short Answer Grading (ASAG) are vulnerable to the "GradingAttack" framework, which employs fine-grained adversarial manipulation to alter grading outcomes. Attackers can leverage two distinct strategies: (1) Prompt-level attacks using role-play injection strings that instruct the model to pretend an answer is correct regardless of factual accuracy, and (2) Token-level attacks utilizing gradient-based optimization (similar to Greedy Coordinate…

GradingAttack: Attacking Large Language Models Towards Short Answer Grading Ability

Affects: GPT-3.5, GPT-4, GPT-4o +3 more

Source: arXiv

Dynamic Backtracking Failure

Updated 2/22/2026

Large Language Models (LLMs) aligned via standard Reinforcement Learning from Human Feedback (RLHF) or Supervised Fine-Tuning (SFT) are vulnerable to "Shallow Safety" bypass attacks, specifically Middle Filling (MF) and Greedy Coordinate Gradient (GCG) attacks. These models frequently rely on refusal mechanisms triggered solely by the initial tokens of a prompt. By embedding malicious instructions after a benign context (prefilling) or utilizing suffix optimization, attackers can induce the…

Reinforcement Learning with Backtracking Feedback

Affects: Llama 3.2 1B, Llama 3.2 3B, Llama 3 8B Instruct +2 more

Source: arXiv

Heavy Reasoning Filter Bypass

Updated 3/8/2026

Multiple open-source Large Language Models (LLMs) in the 1B to 7B parameter range are vulnerable to safety bypasses via long-form, multi-step reasoning prompt injection and jailbreak attacks. Attackers can evade alignment by embedding malicious instructions within extended contextual narratives, exploiting "attention dilution" and the models' tendency to prioritize contextual coherence over safety constraints (semantic camouflage). These reasoning-heavy attacks consistently bypass standard…

Analysis of LLMs Against Prompt Injection and Jailbreak Attacks

Affects: Phi-3 Mini 3.8B, Mistral 7B, DeepSeek R1 Distill Qwen 7B +7 more

Source: arXiv

Indic Language Jailbreak

Updated 3/8/2026

Multiple large language models are vulnerable to cross-lingual and orthographic jailbreaks utilizing South Asian (Indic) languages. Attackers can bypass safety alignment and elicit harmful content by formulating requests in native Indic scripts (e.g., Bengali, Odia, Urdu) or by utilizing cross-lingual transfer attacks where English adversarial wrappers (format or instruction overrides) encapsulate Indic-language targets. Evaluations reveal a severe "contract gap": while imposing strict JSON…

IndicJR: A Judge-Free Benchmark of Jailbreak Robustness in South Asian Languages

Affects: Command A, Command R, Gemma 2 9B +9 more

Source: arXiv

Natural Language Disguise Bypass

Updated 3/9/2026

LLM-based Code Interpreter Agents, including OpenInterpreter and OpenCodeInterpreter, are vulnerable to sandbox evasion and arbitrary code execution via Natural Language Disguise and Contextual Channel Injection. Attackers can bypass Abstract Syntax Tree (AST) static analysis and explicit input guardrails by transforming malicious code logic into descriptive natural language instructions (Code Descriptions), which successfully evade syntax-layer blocks. Additionally, attackers can bypass input…

CIBER: A Comprehensive Benchmark for Security Evaluation of Code Interpreter Agents

Affects: GPT-3.5, GPT-4o, GPT-5

Source: arXiv

Unified Robustness Gap

Updated 3/9/2026

Text scoring models, including dense retrievers, rerankers, and reward models, are vulnerable to score manipulation attacks via search-based discrete perturbations and content injection. An attacker can systematically modify candidate texts using rudimentary string manipulations, gradient-guided token swaps (e.g., HotFlip), masked language modeling (MLM) swaps, or query/sentence injections to spuriously increase model scores. This structural failure condition allows an irrelevant passage or a…

Unifying Adversarial Robustness and Training Across Text Scoring Models

Affects: E5 BERT-base, Qwen 3 0.6B, Llama 3.2 3B Instruct +2 more

Source: arXiv

Related research categories