Imagen de Portada

StructBreak: Structural Cognitive Overload-Induced Safety Failures in MLLMs

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Luo, Yang, Liu, Xinran, Ji, Tiantian, Yin, Zhiyi, Peng, Lingyun, Li, Shuyu
Formato:	Preprint
Publicado:	2026
Materias:	Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

ShadowMerge: A Novel Poisoning Attack on Graph-Based Agent Memory via Relation-Channel Conflicts
por: Luo, Yang, et al.
Publicado: (2026)

Breaking Bad Molecules: Are MLLMs Ready for Structure-Level Molecular Detoxification?
por: Lin, Fei, et al.
Publicado: (2025)

GuardAD: Safeguarding Autonomous Driving MLLMs via Markovian Safety Logic
por: Zhang, Tianyuan, et al.
Publicado: (2026)

StructVRM: Aligning Multimodal Reasoning with Structured and Verifiable Reward Models
por: Zhang, Xiangxiang, et al.
Publicado: (2025)

CSR-Bench: A Benchmark for Evaluating the Cross-modal Safety and Reliability of MLLMs
por: Liu, Yuxuan, et al.
Publicado: (2026)

On-Policy Self-Evolution via Failure Trajectories for Agentic Safety Alignment
por: Yin, Bo, et al.
Publicado: (2026)

Mitigating Societal Cognitive Overload in the Age of AI: Challenges and Directions
por: Lahlou, Salem
Publicado: (2025)

The Forgotten Shield: Safety Grafting in Parameter-Space for Medical MLLMs
por: Zhao, Jiale, et al.
Publicado: (2025)

StructEval: Benchmarking LLMs' Capabilities to Generate Structural Outputs
por: Yang, Jialin, et al.
Publicado: (2025)

Probing Implicit Bias in Semi-gradient Q-learning: Visualizing the Effective Loss Landscapes via the Fokker--Planck Equation
por: Yin, Shuyu, et al.
Publicado: (2024)

Struct-X: Enhancing Large Language Models Reasoning with Structured Data
por: Tan, Xiaoyu, et al.
Publicado: (2024)

StructSynth: Leveraging LLMs for Structure-Aware Tabular Data Synthesis in Low-Data Regimes
por: Liu, Siyi, et al.
Publicado: (2025)

OmniStruct: Universal Text-to-Structure Generation across Diverse Schemas
por: Huang, James Y., et al.
Publicado: (2025)

Explaining and Breaking the Safety-Helpfulness Ceiling via Preference Dimensional Expansion
por: Huang, ShiYing, et al.
Publicado: (2026)

Pragma-VL: Towards a Pragmatic Arbitration of Safety and Helpfulness in MLLMs
por: Wen, Ming, et al.
Publicado: (2026)

VeriStruct: AI-assisted Automated Verification of Data-Structure Modules in Verus
por: Sun, Chuyue, et al.
Publicado: (2025)

AESOP: Adversarial Execution-path Selection to Overload Deep Learning Pipelines
por: Li, Tingxi, et al.
Publicado: (2026)

StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding
por: Lin, Junming, et al.
Publicado: (2024)

FairReason: Balancing Reasoning and Social Bias in MLLMs
por: Pan, Zhenyu, et al.
Publicado: (2025)

StructRAG: Boosting Knowledge Intensive Reasoning of LLMs via Inference-time Hybrid Information Structurization
por: Li, Zhuoqun, et al.
Publicado: (2024)

A priori Estimates for Deep Residual Network in Continuous-time Reinforcement Learning
por: Yin, Shuyu, et al.
Publicado: (2024)

Low-Resource Safety Failures Are Action Failures, Not Representation Failures
por: Aziz, Rashad, et al.
Publicado: (2026)

FlexStructRAG: Flexible Structure-Aware Multi-Granular Relational Retrieval for RAG
por: Chen, Mengzhu, et al.
Publicado: (2026)

StructXLIP: Enhancing Vision-language Models with Multimodal Structural Cues
por: Ruan, Zanxi, et al.
Publicado: (2026)

StructLens: A Structural Lens for Language Models via Maximum Spanning Trees
por: Sakajo, Haruki, et al.
Publicado: (2026)

Image2Struct: Benchmarking Structure Extraction for Vision-Language Models
por: Roberts, Josselin Somerville, et al.
Publicado: (2024)

RL-Struct: A Lightweight Reinforcement Learning Framework for Reliable Structured Output in LLMs
por: Hu, Ruike, et al.
Publicado: (2025)

Enhancing Cognitive Diagnosis by Modeling Learner Cognitive Structure State
por: Chen, Zhifu, et al.
Publicado: (2024)

Diffusion-CAM: Faithful Visual Explanations for dMLLMs
por: Zuo, Haomin, et al.
Publicado: (2026)

StructRL: Recovering Dynamic Programming Structure from Learning Dynamics in Distributional Reinforcement Learning
por: Nowak, Ivo
Publicado: (2026)

StructPrune: Structured Global Pruning asymptotics with $\mathcal{O}(\sqrt{N})$ GPU Memory
por: Song, Xinyuan, et al.
Publicado: (2025)

OmniVideoBench: Towards Audio-Visual Understanding Evaluation for Omni MLLMs
por: Li, Caorui, et al.
Publicado: (2025)

The Geometry of Alignment Collapse: When Fine-Tuning Breaks Safety
por: Springer, Max, et al.
Publicado: (2026)

When Context Flips, Safety Breaks: Diagnosing Brittle Safety in Aligned Language Models
por: Choi, Dasol, et al.
Publicado: (2026)

StructEval: Deepen and Broaden Large Language Model Assessment via Structured Evaluation
por: Cao, Boxi, et al.
Publicado: (2024)

StructMem: Structured Memory for Long-Horizon Behavior in LLMs
por: Xu, Buqiang, et al.
Publicado: (2026)

Affordance Benchmark for MLLMs
por: Wang, Junying, et al.
Publicado: (2025)

Redundancy Principles for MLLMs Benchmarks
por: Zhang, Zicheng, et al.
Publicado: (2025)

Breaking the Martingale Curse: Multi-Agent Debate via Asymmetric Cognitive Potential Energy
por: Liu, Yuhan, et al.
Publicado: (2026)

From Specific-MLLMs to Omni-MLLMs: A Survey on MLLMs Aligned with Multi-modalities
por: Jiang, Shixin, et al.
Publicado: (2024)