Image de couverture de livre

The Missing Half: Unveiling Training-time Implicit Safety Risks Beyond Deployment

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Zhang, Zhexin, Lu, Yida, Fang, Junfeng, Yang, Junxiao, Cui, Shiyao, Zhou, Hao, Meng, Fandong, Zhou, Jie, Wang, Hongning, Huang, Minlie, Chua, Tat-Seng
Format:	Preprint
Publié:	2026
Sujets:	Computation and Language Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Agent-SafetyBench: Evaluating the Safety of LLM Agents
par: Zhang, Zhexin, et autres
Publié: (2024)

Guiding not Forcing: Enhancing the Transferability of Jailbreaking Attacks on LLMs via Removing Superfluous Constraints
par: Yang, Junxiao, et autres
Publié: (2025)

From Theft to Bomb-Making: The Ripple Effect of Unlearning in Defending Against Jailbreak Attacks
par: Zhang, Zhexin, et autres
Publié: (2024)

Be Careful When Fine-tuning On Open-Source LLMs: Your Fine-tuning Data Could Be Secretly Stolen!
par: Zhang, Zhexin, et autres
Publié: (2025)

ShieldVLM: Safeguarding the Multimodal Implicit Toxicity via Deliberative Reasoning with LVLMs
par: Cui, Shiyao, et autres
Publié: (2025)

LongSafety: Evaluating Long-Context Safety of Large Language Models
par: Lu, Yida, et autres
Publié: (2025)

When Smiley Turns Hostile: Interpreting How Emojis Trigger LLMs' Toxicity
par: Cui, Shiyao, et autres
Publié: (2025)

Defending Large Language Models Against Jailbreaking Attacks Through Goal Prioritization
par: Zhang, Zhexin, et autres
Publié: (2023)

LASA: Language-Agnostic Semantic Alignment at the Semantic Bottleneck for LLM Safety
par: Yang, Junxiao, et autres
Publié: (2026)

MiniPLM: Knowledge Distillation for Pre-Training Language Models
par: Gu, Yuxian, et autres
Publié: (2024)

How Should We Enhance the Safety of Large Reasoning Models: An Empirical Study
par: Zhang, Zhexin, et autres
Publié: (2025)

AISafetyLab: A Comprehensive Framework for AI Safety Evaluation and Improvement
par: Zhang, Zhexin, et autres
Publié: (2025)

We Should Identify and Mitigate Third-Party Safety Risks in MCP-Powered Agent Systems
par: Fang, Junfeng, et autres
Publié: (2025)

Beyond Next Token Prediction: Patch-Level Training for Large Language Models
par: Shao, Chenze, et autres
Publié: (2024)

Large Language Models Are Not Robust Multiple Choice Selectors
par: Zheng, Chujie, et autres
Publié: (2023)

DRAFT: Task Decoupled Latent Reasoning for Agent Safety
par: Wang, Lin, et autres
Publié: (2026)

Risky-Bench: Probing Agentic Safety Risks under Real-World Deployment
par: Zheng, Jingnan, et autres
Publié: (2026)

BARREL: Boundary-Aware Reasoning for Factual and Reliable LRMs
par: Yang, Junxiao, et autres
Publié: (2025)

ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors
par: Zhang, Zhexin, et autres
Publié: (2024)

The Side Effects of Being Smart: Safety Risks in MLLMs' Multi-Image Reasoning
par: Chen, Renmiao, et autres
Publié: (2026)

SMSP: A Plug-and-Play Strategy of Multi-Scale Perception for MLLMs to Perceive Visual Illusions
par: Tu, Jinzhe, et autres
Publié: (2026)

3D-TAFS: A Training-free Framework for 3D Affordance Segmentation
par: Chu, Meng, et autres
Publié: (2024)

SafeMLRM: Demystifying Safety in Multi-modal Large Reasoning Models
par: Fang, Junfeng, et autres
Publié: (2025)

Understanding Long Videos via LLM-Powered Entity Relation Graphs
par: Chu, Meng, et autres
Publié: (2025)

SafeNeuron: Neuron-Level Safety Alignment for Large Language Models
par: Wang, Zhaoxin, et autres
Publié: (2026)

JPS: Jailbreak Multimodal Large Language Models with Collaborative Visual Perturbation and Textual Steering
par: Chen, Renmiao, et autres
Publié: (2025)

Compose Your Aesthetics: Empowering Text-to-Image Models with the Principles of Art
par: Jin, Zhe, et autres
Publié: (2025)

On the Implicit Reward Overfitting and the Low-rank Dynamics in RLVR
par: Ye, Hao, et autres
Publié: (2026)

NExT-Guard: Training-Free Streaming Safeguard without Token-Level Labels
par: Fang, Junfeng, et autres
Publié: (2026)

Turing Patterns for Multimedia: Reaction-Diffusion Multi-Modal Fusion for Language-Guided Video Moment Retrieval
par: Fang, Xiang, et autres
Publié: (2026)

Zero-1-to-A: Zero-Shot One Image to Animatable Head Avatars Using Video Diffusion
par: Zhou, Zhenglin, et autres
Publié: (2025)

The Superalignment of Superhuman Intelligence with Large Language Models
par: Huang, Minlie, et autres
Publié: (2024)

CSCD-NS: a Chinese Spelling Check Dataset for Native Speakers
par: Hu, Yong, et autres
Publié: (2022)

Accelerating Inference in Large Language Models with a Unified Layer Skipping Strategy
par: Liu, Yijin, et autres
Publié: (2024)

Continuous Visual Autoregressive Generation via Score Maximization
par: Shao, Chenze, et autres
Publié: (2025)

DeepTrans: Deep Reasoning Translation via Reinforcement Learning
par: Wang, Jiaan, et autres
Publié: (2025)

THOR-MoE: Hierarchical Task-Guided and Context-Responsive Routing for Neural Machine Translation
par: Liang, Yunlong, et autres
Publié: (2025)

ExTrans: Multilingual Deep Reasoning Translation via Exemplar-Enhanced Reinforcement Learning
par: Wang, Jiaan, et autres
Publié: (2025)

Figure It Out: Improve the Frontier of Reasoning with Executable Visual States
par: Chen, Meiqi, et autres
Publié: (2025)

Efficient Covariance Estimation for Sparsified Functional Data
par: Zheng, Sijie, et autres
Publié: (2025)