This paper introduces Security Cube, a comprehensive, multi-dimensional framework for evaluating LLM robustness against jailbreak attacks, providing a systematic taxonomy and benchmark analysis of exi…

The paper proposes MemoAttack, a memory-driven black-box jailbreak framework that systematically models, evolves, and selects attack experiences to significantly enhance LLM jailbreaking success rates…

The paper introduces MT-JailBench, a modular framework for evaluating multi-turn jailbreaks, demonstrating that controlling experimental components like prompt generation and resource budgets is cruci…

OrchJail introduces an orchestration-guided fuzzing framework to systematically jailbreak tool-calling text-to-image agents by exploiting unsafe multi-step tool-orchestration patterns.

This paper systematically analyzes the interaction of multiple weak jailbreak attacks (mutators) applied sequentially to LLMs, finding that most combinations fail due to destructive interference, reve…

The paper proposes an unsupervised bi-level adversarial training framework to enhance LLM safety steering, achieving strong zero-shot defense against unseen and evolving jailbreak prompts.

D-Judge introduces a semantics-preserving output rewriting defense that disrupts multi-turn jailbreak attacks by misaligning the feedback signal used by an attacker's judge model.

This paper addresses the lack of systematic infrastructure for evaluating jailbreak attacks by introducing a large-scale dataset, an automated generation method, and a continuous evaluation metric tha…

The paper investigates how different methods of jailbreaking large language models (SFT, RLVR, and abliteration) lead to vastly different behavioral and mechanistic failures, even when all methods ach…

The paper proposes TRACE, a novel agentic jailbreaking framework that successfully bypasses safety mechanisms of advanced LLM agents by decomposing malicious tasks and disguising harmful subtasks with…

Membrane introduces a self-evolving guardrail using Contrastive Safety Memory (CSM) that generalizes across topical jailbreak variants, achieving superior safety performance while minimizing benign re…

The paper introduces Persona Attack, a novel memory injection jailbreak method that demonstrates that accumulating instructions in the model's context window can override internal safety alignments, a…

The paper introduces Persona Attack, a novel memory injection jailbreak method that demonstrates how accumulating instructions in the model's context window can override internal safety alignments, ac…

The paper conducts an interpretability-driven safety audit of eight state-of-the-art LLMs, demonstrating that while interpretability-based steering is a powerful auditing tool, model robustness varies…

The paper proposes mitigating the progressive degradation of safety in language models caused by many-shot jailbreak attacks by appending a single, fixed safety demonstration at inference time.

The paper introduces Incremental Completion Decomposition (ICD), a novel jailbreak strategy that successfully bypasses LLM safety mechanisms by eliciting malicious content through a sequence of single…

This paper introduces a novel backdoor attack (ACB) against Reinforcement Learning with Verifiable Rewards (RLVR), demonstrating that poisoning the training data can implant a backdoor that significan…

The paper introduces EvoJail, an automated multi-objective evolutionary framework that systematically discovers diverse and effective long-tail jailbreak attacks against LLMs by optimizing for attack…

The paper proposes GUARD-SLM, a token activation-based defense mechanism, to enhance the robustness of Small Language Models (SLMs) against various jailbreak attacks by analyzing and filtering malicio…

NeuroArmor is a white-box runtime defense that uses prompt-specific safe variants to selectively detect and mitigate jailbreak attacks, significantly reducing attack success rates while maintaining a…