{"total":63,"rows":[{"id":"2406.05946","title":"Safety Alignment Should Be Made More Than Just a Few Tokens Deep","created":"2024-06-10","cited_by_count":472,"area":"alignment"},{"id":"2408.17003","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","created":"2024-08-30","cited_by_count":134,"area":"alignment"},{"id":"2502.17420","title":"The Geometry of Refusal in Large Language Models: Concept Cones and Representational Independence","created":"2025-02-24","cited_by_count":115,"area":"alignment"},{"id":"2405.18641","title":"Lisa: Lazy Safety Alignment for Large Language Models against Harmful Fine-tuning Attack","created":"2024-05-28","cited_by_count":92,"area":"alignment"},{"id":"2409.18169","title":"Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey","created":"2024-09-26","cited_by_count":90,"area":"alignment"},{"id":"2408.09600","title":"Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning","created":"2024-08-18","cited_by_count":72,"area":"alignment"},{"id":"2406.05954","title":"Aligning Large Language Models with Representation Editing: A Control Perspective","created":"2024-06-10","cited_by_count":64,"area":"alignment"},{"id":"2406.14144","title":"Towards Understanding Safety Alignment: A Mechanistic Perspective from Safety Neurons","created":"2024-06-20","cited_by_count":53,"area":"alignment"},{"id":"2410.09760","title":"Targeted Vaccine: Safety Alignment for Large Language Models against Harmful Fine-Tuning via Layer-wise Perturbation","created":"2024-10-13","cited_by_count":37,"area":"alignment"},{"id":"2412.12497","title":"NLSR: Neuron-Level Safety Realignment of Large Language Models Against Harmful Fine-Tuning","created":"2024-12-17","cited_by_count":37,"area":"alignment"},{"id":"2405.13967","title":"Model Editing as a Robust and Denoised variant of DPO: A Case Study on Toxicity","created":"2024-05-22","cited_by_count":35,"area":"alignment"},{"id":"2406.16235","title":"Preference Tuning For Toxicity Mitigation Generalizes Across Languages","created":"2024-06-23","cited_by_count":31,"area":"alignment"},{"id":"2405.16229","title":"No Two Devils Alike: Unveiling Distinct Mechanisms of Fine-tuning Attacks","created":"2024-05-25","cited_by_count":29,"area":"alignment"},{"id":"2503.17239","title":"SafeMERGE: Preserving Safety Alignment in Fine-Tuned Large Language Models via Selective Layer-Wise Model Merging","created":"2025-03-21","cited_by_count":25,"area":"alignment"},{"id":"2408.01420","title":"Mission Impossible: A Statistical Perspective on Jailbreaking LLMs","created":"2024-08-02","cited_by_count":24,"area":"alignment"},{"id":"2412.11041","title":"Separate the Wheat from the Chaff: A Post-Hoc Approach to Safety Re-Alignment for Fine-Tuned Language Models","created":"2024-12-15","cited_by_count":20,"area":"alignment"},{"id":"2506.05346","title":"Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets","created":"2025-06-05","cited_by_count":20,"area":"alignment"},{"id":"2405.13820","title":"Towards Comprehensive Post Safety Alignment of Large Language Models via Safety Patching","created":"2024-05-22","cited_by_count":17,"area":"alignment"},{"id":"2410.10862","title":"Superficial Safety Alignment Hypothesis","created":"2024-10-07","cited_by_count":14,"area":"alignment"},{"id":"2502.01208","title":"On Almost Surely Safe Alignment of Large Language Models at Inference-Time","created":"2025-02-03","cited_by_count":14,"area":"alignment"},{"id":"2505.18588","title":"Safety Alignment via Constrained Knowledge Unlearning","created":"2025-05-24","cited_by_count":14,"area":"alignment"},{"id":"2507.21182","title":"SDD: Self-Degraded Defense against Malicious Fine-tuning","created":"2025-07-27","cited_by_count":13,"area":"alignment"},{"id":"2508.07172","title":"Gradient Surgery for Safe LLM Fine-Tuning","created":"2025-08-10","cited_by_count":11,"area":"alignment"},{"id":"2506.18931","title":"Safe Pruning LoRA: Robust Distance-Guided Pruning for Safety Alignment in Adaptation of LLMs","created":"2025-06-21","cited_by_count":10,"area":"alignment"},{"id":"2510.01243","title":"Detoxifying Large Language Models via Autoregressive Reward Guided Representation Editing","created":"2025-09-24","cited_by_count":10,"area":"alignment"},{"id":"2505.17072","title":"Safety Alignment Can Be Not Superficial With Explicit Safety Signals","created":"2025-05-19","cited_by_count":9,"area":"alignment"},{"id":"2508.05775","title":"Guardians and Offenders: A Survey on Harmful Content Generation and Safety Mitigation of LLM","created":"2025-08-07","cited_by_count":9,"area":"alignment"},{"id":"2510.14301","title":"A Guardrail for Safety Preservation: When Safety-Sensitive Subspace Meets Harmful-Resistant Null-Space","created":"2025-10-16","cited_by_count":9,"area":"alignment"},{"id":"2601.01887","title":"Safety at One Shot: Patching Fine-Tuned LLMs with A Single Instance","created":"2026-01-05","cited_by_count":8,"area":"alignment"},{"id":"2602.00038","title":"LSSF: Safety Alignment for Large Language Models through Low-Rank Safety Subspace Fusion","created":"2026-01-19","cited_by_count":8,"area":"alignment"},{"id":"2503.19041","title":"LookAhead Tuning: Safer Language Models via Partial Answer Previews","created":"2025-03-24","cited_by_count":7,"area":"alignment"},{"id":"2502.13946","title":"Why Safeguarded Ships Run Aground? Aligned Large Language Models' Safety Mechanisms Tend to Be Anchored in The Template Region","created":"2025-02-19","cited_by_count":6,"area":"alignment"},{"id":"2505.14185","title":"Safety Subspaces are Not Linearly Distinct: A Fine-Tuning Case Study","created":"2025-05-20","cited_by_count":6,"area":"alignment"},{"id":"2411.06424","title":"How Does DPO Reduce Toxicity? A Mechanistic Neuron-Level Analysis","created":"2024-11-10","cited_by_count":5,"area":"alignment"},{"id":"2508.19697","title":"Safety Alignment Should Be Made More Than Just A Few Attention Heads","created":"2025-08-27","cited_by_count":5,"area":"alignment"},{"id":"2508.20766","title":"Turning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety Injection","created":"2025-08-28","cited_by_count":5,"area":"alignment"},{"id":"2511.06852","title":"Differentiated Directional Intervention A Framework for Evading LLM Safety Alignment","created":"2025-11-10","cited_by_count":5,"area":"alignment"},{"id":"2512.23260","title":"Interpretable Safety Alignment via SAE-Constructed Low-Rank Subspace Adaptation","created":"2025-12-29","cited_by_count":5,"area":"alignment"},{"id":"2504.09895","title":"Learning from Reference Answers: Versatile Language Model Alignment without Binary Human Preference Data","created":"2025-04-14","cited_by_count":4,"area":"alignment"},{"id":"2508.09473","title":"NeuronTune: Fine-Grained Neuron Modulation for Balanced Safety-Utility Alignment in LLMs","created":"2025-08-13","cited_by_count":4,"area":"alignment"},{"id":"2602.07340","title":"Revisiting Robustness for LLM Safety Alignment via Selective Geometry Control","created":"2026-02-07","cited_by_count":4,"area":"alignment"},{"id":"2604.12384","title":"Preventing Safety Drift in Large Language Models via Coupled Weight and Activation Constraints","created":"2026-04-14","cited_by_count":4,"area":"alignment"},{"id":"2406.01931","title":"Dishonesty in Helpful and Harmless Alignment","created":"2024-06-04","cited_by_count":3,"area":"alignment"},{"id":"2602.04448","title":"RASA: Routing-Aware Safety Alignment for Mixture-of-Experts Models","created":"2026-02-04","cited_by_count":3,"area":"alignment"},{"id":"2605.08513","title":"A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models","created":"2026-05-08","cited_by_count":3,"area":"alignment"},{"id":"2507.00239","title":"Linearly Decoding Refused Knowledge in Aligned Language Models","created":"2025-06-30","cited_by_count":2,"area":"alignment"},{"id":"2602.16977","title":"Fail-Closed Alignment for Large Language Models","created":"2026-02-19","cited_by_count":2,"area":"alignment"},{"id":"2603.14355","title":"Exposing Long-Tail Safety Failures in Large Language Models through Efficient Diverse Response Sampling","created":"2026-03-15","cited_by_count":2,"area":"alignment"},{"id":"2605.18309","title":"Alignment Dynamics in LLM Fine-Tuning","created":"2026-05-18","cited_by_count":2,"area":"alignment"},{"id":"2506.00676","title":"SafeTuneBed: A Toolkit for Benchmarking LLM Safety Alignment in Fine-Tuning","created":"2025-05-31","cited_by_count":1,"area":"alignment"},{"id":"2509.03985","title":"NeuroBreak: Unveil Internal Jailbreak Mechanisms in Large Language Models","created":"2025-09-04","cited_by_count":1,"area":"alignment"},{"id":"2509.12221","title":"MEUV: Achieving Fine-Grained Capability Activation in Large Language Models via Mutually Exclusive Unlock Vectors","created":"2025-09-04","cited_by_count":0,"area":"alignment"},{"id":"2510.09004","title":"Decoupling Safety into Orthogonal Subspace: Cost-Efficient and Performance-Preserving Alignment for Large Language Models","created":"2025-10-10","cited_by_count":0,"area":"alignment"},{"id":"2601.03005","title":"JPU: Bridging Jailbreak Defense and Unlearning via On-Policy Path Rectification","created":"2026-01-06","cited_by_count":0,"area":"alignment"},{"id":"2605.04572","title":"From Parameter Dynamics to Risk Scoring : Quantifying Sample-Level Safety Degradation in LLM Fine-tuning","created":"2026-05-06","cited_by_count":0,"area":"alignment"},{"id":"2605.17971","title":"Babel: Jailbreaking Safety Attention via Obfuscation Distribution Optimized Sampling","created":"2026-05-18","cited_by_count":0,"area":"alignment"},{"id":"2605.20654","title":"REFLECTOR: Internalizing Step-wise Reflection against Indirect Jailbreak","created":"2026-05-20","cited_by_count":0,"area":"alignment"},{"id":"2607.11475","title":"HyperSafe: Inference-Time Safety Recovery for Fine-Tuned Language Models","created":"2026-07-13","cited_by_count":0,"area":"alignment"},{"id":"2608.05045","title":"Gradient Immunity: Null-Space Resistance to Malicious Fine-Tuning","created":"2026-08-05","cited_by_count":0,"area":"alignment"},{"id":"2608.08542","title":"When Skills Meet Safety: Benchmarking and Characterizing the Adaptive Jailbreak Robustness of Skill-Merged LLMs","created":"2026-08-09","cited_by_count":0,"area":"alignment"},{"id":"2608.11583","title":"Localizing Safety Alignment: MLP Layers and Mid-Network Blocks Encode Refusal Behavior in Large Language Models","created":"2026-08-12","cited_by_count":0,"area":"alignment"},{"id":"2609.01455","title":"When Safety Routing Breaks: Understanding Alignment Fragility under Benign Fine-Tuning","created":"2026-09-01","cited_by_count":0,"area":"alignment"},{"id":"2609.06934","title":"The Geometry of Refusal: Why Post-Hoc Safety Is Fragile and Pretraining-Time Safety Persists","created":"2026-09-07","cited_by_count":0,"area":"alignment"}],"next_cursor":null}