🔗 Who cites «Safety Alignment Should Be Made More Than Just a Few Tokens Deep»

93 papers of our library cite it (of the 149,439 whose bibliographies we read) · 35 with a page here, the most cited first · ← back to «Safety Alignment Should Be Made More Than Just a Few Tokens Deep»

1. Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive AttacksarXiv:2404.02151Maksym Andriushchenko, Francesco Croce, Nicolas Flammarion2024 · 118 citations
2. The Multilingual Alignment Prism: Aligning Global and Local Preferences to Reduce HarmConference on Empirical Methods in Natural Language Processing (EMNLP)Aakanksha, Arash Ahmadian, Beyza Ermis, et al.2024 · 49 citations
3. Reasoning-to-Defend: Safety-Aware Reasoning Can Defend Large Language Models from JailbreakingConference on Empirical Methods in Natural Language Processing (EMNLP)Junda Zhu, Lingyong Yan, Shuaiqiang Wang, et al.2025 · 21 citations
4. Refuse Whenever You Feel Unsafe: Improving Safety in LLMs via Decoupled Refusal TrainingAnnual Meeting of the Association for Computational Linguistics (ACL)Youliang Yuan, Wenxiang Jiao, Wenxuan Wang, et al.2025 · 19 citations
5. Distilling LLM Agent into Small Models with Retrieval and Code ToolsarXiv:2505.17612Minki Kang, Jongwon Jeong, Seanie Lee, et al.2025 · 11 citations
6. Layer-Aware Representation Filtering: Purifying Finetuning Data to Preserve LLM Safety AlignmentConference on Empirical Methods in Natural Language Processing (EMNLP)Hao Li, Lijun Li, Zhenghao Lu, et al.2025 · 11 citations
7. Backdoors in RLVR: Jailbreak Backdoors in LLMs From Verifiable RewardAnnual Meeting of the Association for Computational Linguistics (ACL)Weiyang Guo, Zesheng Shi, Zeen Zhu, et al.2026 · 10 citations
8. Rethinking Jailbreak Detection of Large Vision Language Models with Representational Contrastive ScoringAnnual Meeting of the Association for Computational Linguistics (ACL)Peichun Hua, Hao Li, Shanghao Shi, et al.2026 · 10 citations
9. SafeKey: Amplifying Aha-Moment Insights for Safety ReasoningConference on Empirical Methods in Natural Language Processing (EMNLP)Kaiwen Zhou, Xuandong Zhao, Jayanth Srinivasa, et al.2025 · 9 citations
10. Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning DatasetsAnnual Meeting of the Association for Computational Linguistics (ACL)Lei Hsiung, Tianyu Pang, Yung-Chen Tang, et al.2026 · 4 citations
11. SDD: Self-Degraded Defense against Malicious Fine-tuningAnnual Meeting of the Association for Computational Linguistics (ACL)ZiXuan Chen, Weikai Lu, Xin Lin, et al.2025 · 4 citations
12. CTRAP: Embedding Collapse Trap to Safeguard Large Language Models from Harmful Fine-TuningAnnual Meeting of the Association for Computational Linguistics (ACL)Biao Yi, Tiansheng Huang, Baolei Zhang, et al.2026 · 3 citations
13. Universal Jailbreak Suffixes Are Strong Attention HijackersTransactions of the Association for Computational Linguistics (TACL)Matan Ben-Tov, Mor Geva, Mahmood Sharif2026 · 3 citations
14. Activation-Guided Local Editing for Jailbreaking AttacksAnnual Meeting of the Association for Computational Linguistics (ACL)Jiecong Wang, Haoran Li, Hao Peng, et al.2026 · 1 citation
15. Beyond Surface-Level Detection: Towards Cognitive-Driven Defense Against Jailbreak Attacks via Meta-Operations ReasoningAnnual Meeting of the Association for Computational Linguistics (ACL)Rui Pu, Chaozhuo Li, Rui Ha, et al.2026 · 1 citation
16. Reasoning Up the Instruction Ladder for Controllable Language ModelsFindings of the Association for Computational LinguisticsZishuo Zheng, Vidhisha Balachandran, Chan Young Park, et al.2026 · 1 citation
17. Can Small-Scale Data Poisoning Exacerbate Dialect-Linked Biases in Large Language Models?arXiv:2507.19195Chaymaa Abbas, Mariette Awad, Razane Tajeddine2025 · 1 citation
18. HAMSA: Hijacking Aligned Compact Models via Stealthy AutomationarXiv:2508.16484Alexey Krylov, Iskander Vagizov, Dmitrii Korzh, et al.2025 · 1 citation
19. Deactivating Refusal Triggers: Understanding and Mitigating Overrefusal in Safety AlignmentWorkshop on Trustworthy NLPZhiyu Xue, Zimo Qi, Guangliang Liu, et al.2026
20. From Narrow Unlearning to Emergent Misalignment in LLMsAnnual Meeting of the Association for Computational Linguistics (ACL)Erum Mushtaq, Anil Ramakrishna, Satyapriya Krishna, et al.2026
21. How Do LLMs and VLMs Understand Viewpoint Rotation Without Vision? An Interpretability StudyAnnual Meeting of the Association for Computational Linguistics (ACL)Zhen Yang, Ping Jian, Zhongbin Guo, et al.2026
22. Let's Put Ourselves in Sally's Shoes: Shoes-of-Others Prefilling Improves Theory of Mind in Large Language ModelsFindings of the Association for Computational LinguisticsKazutoshi Shinoda, Nobukatsu Hojo, Kyosuke Nishida, et al.2026
23. More Thinking, Less Talking: Internalizing Deliberative Safety into LLM ParametersAnnual Meeting of the Association for Computational Linguistics (ACL)Guan Wang, Xuehai Tang, Biyu Zhou, et al.2026
24. OASIS: Mitigating Harmful Fine-tuning Attacks on LLMs via Orthogonal and Adaptive Safety Alignment StrategyAnnual Meeting of the Association for Computational Linguistics (ACL)Jiayu Tang, Guowei Peng, Qiuhao Xie, et al.2026
25. Pruning Unsafe Tickets: A Resource-Efficient Framework for Safer and More Robust LLMsAnnual Meeting of the Association for Computational Linguistics (ACL)Wai Man Si, Mingjie Li, Michael Backes, et al.2026
26. Rejection-to-Acceptance Transition: Model Editing-Based Jailbreak Backdoor Injection Not Limited to Few Output TokensFindings of the Association for Computational LinguisticsShiji Yang, Min Cai, Hao Xiong, et al.2026
27. Resolving the Security-Auditability Dilemma with Auditable Latent Chain-of-Thought AlignmentAnnual Meeting of the Association for Computational Linguistics (ACL)Guan Wang, Biyu Zhou, Xuehai Tang, et al.2026
28. Retrieval-Augmented Defense: Adaptive and Controllable Jailbreak Prevention for Large Language ModelsAnnual Meeting of the Association for Computational Linguistics (ACL)Guangyu Yang, Jinghong Chen, Jingbiao Mei, et al.2026
29. Revisiting Self-Play Preference Optimization: On the Role of Prompt DifficultyFindings of the Association for Computational LinguisticsYao Xiao, Jung-jae Kim, Roy Ka-Wei Lee, et al.2026
30. Safeguarding LLM Fine-tuning via Push-Pull Distributional AlignmentAnnual Meeting of the Association for Computational Linguistics (ACL)Haozhong Wang, Zhuo Li, Yibo Yang, et al.2026
31. StealthGraph: Exposing Domain-Specific Risks in LLMs through Knowledge-Graph-Guided Harmful Prompt GenerationAnnual Meeting of the Association for Computational Linguistics (ACL)Huawei Zheng, Xinqi Jiang, Sen Yang, et al.2026
32. The Art of (Mis)alignment: How Fine-Tuning Methods Effectively Misalign and Realign LLMs in Post-TrainingFindings of the Association for Computational LinguisticsRui Zhang, Hongwei Li, Yun Shen, et al.2026
33. Thinking Twice Makes Large Language Models Safer and More HelpfulFindings of the Association for Computational LinguisticsYutao Mou, Yuxiao Luo, Shikun Zhang, et al.2026
34. TriPlay-RL: Tri-Role Self-Play Reinforcement Learning for LLM Safety AlignmentAnnual Meeting of the Association for Computational Linguistics (ACL)Zhewen Tan, Wenhan Yu, Jianfeng Si, et al.2026
35. When Safety Alignment Fails to Generalize: Probing with Language Game JailbreaksFindings of the Association for Computational LinguisticsZewen Long, Yu Peng, Fangming Dong, et al.2026

58 more papers of our library cite it; they have no page here yet.

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/qi-2024-safety-alignment-should-made-more/cited-by/