Cover Image

Strata-Sword: A Hierarchical Safety Evaluation towards LLMs based on Reasoning Complexity of Jailbreak Instructions

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Zhao, Shiji, Duan, Ranjie, Liu, Jiexi, Jia, Xiaojun, Wang, Fengxiang, Wei, Cheng, Cheng, Ruoxi, Xie, Yong, Liu, Chang, Guo, Qing, Tao, Jialing, Xue, Hui, Wei, Xingxing
Format:	Preprint
Published:	2025
Subjects:	Computers and Society
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Jailbreaking Multimodal Large Language Models via Shuffle Inconsistency
by: Zhao, Shiji, et al.
Published: (2025)

MRJ-Agent: An Effective Jailbreak Agent for Multi-Round Dialogue
by: Wang, Fengxiang, et al.
Published: (2024)

VRSA: Jailbreaking Multimodal Large Language Models through Visual Reasoning Sequential Attack
by: Zhao, Shiji, et al.
Published: (2025)

Improving Adversarial Robust Fairness via Anti-Bias Soft Label Distillation
by: Zhao, Shiji, et al.
Published: (2023)

MoAPT: Mixture of Adversarial Prompt Tuning for Vision-Language Models
by: Zhao, Shiji, et al.
Published: (2025)

Towards Class-wise Fair Adversarial Training via Anti-Bias Soft Label Distillation
by: Zhao, Shiji, et al.
Published: (2025)

Oyster-I: Beyond Refusal -- Constructive Safety Alignment for Responsible Language Models
by: Duan, Ranjie, et al.
Published: (2025)

Inverse Reinforcement Learning with Dynamic Reward Scaling for LLM Alignment
by: Cheng, Ruoxi, et al.
Published: (2025)

PBI-Attack: Prior-Guided Bimodal Interactive Black-Box Jailbreak Attack for Toxicity Maximization
by: Cheng, Ruoxi, et al.
Published: (2024)

Knowledge-Guided Adversarial Training for Infrared Object Detection via Thermal Radiation Modeling
by: Zhao, Shiji, et al.
Published: (2026)

HTS-Attack: Heuristic Token Search for Jailbreaking Text-to-Image Models
by: Gao, Sensen, et al.
Published: (2024)

Mirage in the Eyes: Hallucination Attack on Multi-modal Large Language Models with Only Attention Sink
by: Wang, Yining, et al.
Published: (2025)

A Causal Perspective for Enhancing Jailbreak Attack and Defense
by: Pan, Licheng, et al.
Published: (2026)

MESA: Improving MoE Safety Alignment via Decentralized Expertise
by: Sun, Yitong, et al.
Published: (2026)

Obscure but Effective: Classical Chinese Jailbreak Prompt Optimization via Bio-Inspired Search
by: Huang, Xun, et al.
Published: (2026)

Mitigating Accuracy-Robustness Trade-off via Balanced Multi-Teacher Adversarial Distillation
by: Zhao, Shiji, et al.
Published: (2023)

Revisiting the Trade-off between Accuracy and Robustness via Weight Distribution of Filters
by: Wei, Xingxing, et al.
Published: (2023)

Heuristic-Induced Multimodal Risk Distribution Jailbreak Attack for Multimodal Large Language Models
by: Teng, Ma, et al.
Published: (2024)

The Trojan Example: Jailbreaking LLMs through Template Filling and Unsafety Reasoning
by: Liu, Mingrui, et al.
Published: (2025)

Improving Safety Alignment via Balanced Direct Preference Optimization
by: Zhao, Shiji, et al.
Published: (2026)

Empowering LLMs with Logical Reasoning: A Comprehensive Survey
by: Cheng, Fengxiang, et al.
Published: (2025)

Reasoned Safety Alignment: Ensuring Jailbreak Defense via Answer-Then-Check
by: Cao, Chentao, et al.
Published: (2025)

Reasoning as an Attack Surface: Adaptive Evolutionary CoT Jailbreaks for LLMs
by: Li, Jianan, et al.
Published: (2026)

How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMs
by: Zeng, Yi, et al.
Published: (2024)

GUARD: Guideline Upholding Test through Adaptive Role-play and Jailbreak Diagnostics for LLMs
by: Jin, Haibo, et al.
Published: (2025)

Adaptive Selection of Symbolic Languages for Improving LLM Logical Reasoning
by: Wang, Xiangyu, et al.
Published: (2025)

TUNI: A Textual Unimodal Detector for Identity Inference in CLIP Models
by: Li, Songze, et al.
Published: (2024)

When Lighting Deceives: Exposing Vision-Language Models' Illumination Vulnerability Through Illumination Transformation Attack
by: Liu, Hanqing, et al.
Published: (2025)

NDM: A Noise-driven Detection and Mitigation Framework against Implicit Sexual Intentions in Text-to-Image Generation
by: Sun, Yitong, et al.
Published: (2025)

MathScale: Scaling Instruction Tuning for Mathematical Reasoning
by: Tang, Zhengyang, et al.
Published: (2024)

Towards Arbitrary Motion Completing via Hierarchical Continuous Representation
by: Xu, Chenghao, et al.
Published: (2025)

Iterative Self-Tuning LLMs for Enhanced Jailbreaking Capabilities
by: Sun, Chung-En, et al.
Published: (2024)

Sword and Shield: Uses and Strategies of LLMs in Navigating Disinformation
by: Lim, Gionnieve, et al.
Published: (2025)

STAIR: Improving Safety Alignment with Introspective Reasoning
by: Zhang, Yichi, et al.
Published: (2025)

Adaptive Interviewing for Persona Simulation in LLMs: Evidence-Grounded Reasoning Improves Decision Alignment
by: Su, Ruoxi, et al.
Published: (2026)

Inducing Overthink: Hierarchical Genetic Algorithm-based DoS Attack on Black-Box Large Language Reasoning Models
by: Wang, Shuqiang, et al.
Published: (2026)

PricingLogic: Evaluating LLMs Reasoning on Complex Tourism Pricing Tasks
by: Liu, Yunuo, et al.
Published: (2025)

Adversarial Orthogonal Disentanglement for LVLM Hallucination Mitigation
by: Cheng, Ruoxi, et al.
Published: (2026)

The Path to Reconciling Quality and Safety in Text-to-Image Generation: Dataset, Method, and Evaluation
by: Ruan, Shouwei, et al.
Published: (2025)

Learning to Detect Unknown Jailbreak Attacks in Large Vision-Language Models
by: Liang, Shuang, et al.
Published: (2025)