The paper introduces RASA, a routing‑aware safety alignment framework for Mixture‑of‑Experts (MoE) language models. RASA repairs Safety‑Critical Experts by selectively fine‑tuning only those experts that are disproportionately activated during jailbreak attacks, while keeping routing fixed and enforcing routing consistency with safety‑aligned contexts. Experiments on two MoE architectures show that RASA achieves near‑perfect robustness against diverse jailbreak attacks, generalizes across attacks, reduces over‑refusal, and maintains performance on standard benchmarks such as MMLU, GSM8K, and TruthfulQA.
By Jiacheng Liang, Yuhui Wang, Tanqiu Jiang, Ting Wang
arXiv:2607. 06109v1 Announce Type: cross Abstract: Multi-perturbation adversarial training (MAT) aims to achieve robustness against multiple $\ell_p$ perturbations but suffers from robustness trade-offs between different threats.
By Woo Jae Kim, Kyle Min, Suhyeon Ha, Joonsung Jeon, Sung-eui Yoon
The paper examines the vulnerability of open‑weight large language model safeguards that aim to block harmful usage by assuming new harmful behavior arises only through fine‑tuning. It demonstrates that simple, low‑cost attacks—abliteration and prefilling—can significantly increase attack success rates across multiple benchmarks, and proposes an abliteration‑resistant tuning (ART) method that mitigates these attacks. The study shows that safeguarding defenses must be evaluated against a broader range of attack strategies beyond fine‑tuning.
By Kevin Kuo, Virginia Smith, Chhavi Yadav
NeuronGuard is a fine‑tuning defense for large language models that hardens them against both jailbreak and neuron‑level attacks. It redistributes safety signals across many neurons by identifying safety‑critical ones with per‑layer linear classifiers, enforcing refusal behavior when those neurons are ablated, and applying KL‑divergence regularization for consistency. A randomized gradient projection preserves task performance, and the authors provide a formal guarantee that NeuronGuard lowers the attack success rate upper bound, with experiments showing near‑zero success rates across multiple models and attack strategies.
By Anjun Gao, Yueyang Quan, Yufei Xia, Zhuqing Liu, Minghong Fang
Large language models (LLMs) are vulnerable to backdoor attacks, where hidden triggers induce malicious outputs. Existing defenses generally fall into inference-time detection or training-time mitigation, but face two key limitations.
arXiv:2602. 16835v2 Announce Type: replace-cross Abstract: Safety alignment is essential for the responsible deployment of Large Language Models (LLMs).
By Sasha Behrouzi, Lichao Wu, Mohamadreza Rostami, Ahmad-Reza Sadeghi
arXiv:2606. 11409v1 Announce Type: cross Abstract: Adversarial robustness evaluations of large language models (LLMs) typically report attack success rate (ASR) under fixed query budgets, implicitly treating all attacks as equally costly.
By Malikeh Ehghaghi, Bogl\'arka Ecsedi, Marsha Chechik, Colin Raffel
arXiv:2607. 19894v1 Announce Type: cross Abstract: Large language models (LLMs) are vulnerable to backdoor attacks, where hidden triggers induce malicious outputs.
By Yuxi Li, Zhibo Zhang, Kailong Wang, Xingshuo Han, Ling Shi, Haoyu Wang
arXiv:2606. 15441v1 Announce Type: cross Abstract: Indirect prompt injection attacks hijack LLM-based agents by embedding malicious instructions in third-party data that the agent retrieves during task execution.
By Lipeng He, Yihan Wang, Jiawen Zhang, N. Asokan
Indirect prompt injection attacks hijack LLM-based agents by embedding malicious instructions in third-party data that the agent retrieves during task execution. Existing defenses report near-zero attack success rate on static benchmarks, yet recent adaptive evaluations show that these results collapse once the attacker is allowed to optimize against the deployed defense.
arXiv:2510. 02422v4 Announce Type: replace-cross Abstract: Existing gradient-based jailbreak attacks typically optimize a fixed-length adversarial suffix toward a predefined target response with a static optimization strategy.
By Kedong Xiu, Yunhan Yang, Churui Zeng, Tianhang Zheng, Xinzhe Huang, Di Wang, Puning Zhao, Zhan Qin, Kui Ren
arXiv:2607. 08883v1 Announce Type: new Abstract: Behavioral alignment in large language models often masks fragile internal safety representations.
By Ege \c{C}akar, Hannah Guan, Kayden Kehe