When Skills Meet Safety: Benchmarking and Characterizing the Adaptive Jailbreak Robustness of Skill-Merged LLMs
Yu Ma, Hongli Shi, Jing Li, Xinran Xu, Weiwei Hou
cs.LG
Submitted: 2026-08-09
Updated: 2026-08-11
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks
- Refusal in Language Models Is Mediated by a Single Direction
- Language Models are Homer Simpson! Safety Re-Alignment of Fine-tuned Language Models through Task Arithmetic
- JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models
- Jailbreaking Black Box Large Language Models in Twenty Queries
- Evaluating Large Language Models Trained on Code
- Training Verifiers to Solve Math Word Problems
- Colluding LoRA: A Compositional Vulnerability in LLM Safety Alignment
- SafeMERGE: Preserving Safety Alignment in Fine-Tuned Large Language Models via Selective Layer-Wise Model Merging
- Safeguard Fine-Tuned LLMs Through Pre- and Post-Tuning Model Merging
- Merging Improves Self-Critique Against Jailbreak Attacks
- Model Merging and Safety Alignment: One Bad Model Spoils the Bunch
- MergeBench: A Benchmark for Merging Domain-Specialized LLMs
- Measuring Massive Multitask Language Understanding
- Safe LoRA: the Silver Lining of Reducing Safety Risks when Fine-tuning Large Language Models
- Editing Models with Task Arithmetic
- Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
- LoRA Fine-tuning Efficiently Undoes Safety Training in Llama 2-Chat 70B
- When Safe Models Merge into Danger: Exploiting Latent Vulnerabilities in LLM Fusion
- LED-Merging: Mitigating Safety-Utility Conflicts in Model Merging with Location-Election-Disjoint
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks