{"total":96,"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":"2406.14598","title":"SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal","created":"2024-06-20","cited_by_count":240,"area":"alignment"},{"id":"2402.05162","title":"Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications","created":"2024-02-07","cited_by_count":231,"area":"alignment"},{"id":"2310.15140","title":"AutoDAN: Interpretable Gradient-Based Adversarial Attacks on Large Language Models","created":"2023-10-23","cited_by_count":111,"area":"alignment"},{"id":"2402.19085","title":"Controllable Preference Optimization: Toward Controllable Multi-Objective Alignment","created":"2024-02-29","cited_by_count":107,"area":"alignment"},{"id":"2402.09063","title":"Soft Prompt Threats: Attacking Safety Alignment and Unlearning in Open-Source LLMs through the Embedding Space","created":"2024-02-14","cited_by_count":91,"area":"alignment"},{"id":"2401.11206","title":"InferAligner: Inference-Time Alignment for Harmlessness through Cross-Model Guidance","created":"2024-01-20","cited_by_count":83,"area":"alignment"},{"id":"2402.12329","title":"Query-Based Adversarial Prompt Generation","created":"2024-02-19","cited_by_count":62,"area":"alignment"},{"id":"2404.16369","title":"Don't Say No: Jailbreaking LLM by Suppressing Refusal","created":"2024-04-25","cited_by_count":56,"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":"2408.11491","title":"SCANS: Mitigating the Exaggerated Safety for LLMs via Safety-Conscious Activation Steering","created":"2024-08-21","cited_by_count":50,"area":"alignment"},{"id":"2403.07230","title":"Curry-DPO: Enhancing Alignment using Curriculum Learning & Ranked Preferences","created":"2024-03-12","cited_by_count":39,"area":"alignment"},{"id":"2504.09466","title":"AdaSteer: Your Aligned LLM is Inherently an Adaptive Jailbreak Defender","created":"2025-04-13","cited_by_count":38,"area":"alignment"},{"id":"2406.11801","title":"Safety Arithmetic: A Framework for Test-time Safety Alignment of Language Models by Steering Parameters and Activations","created":"2024-06-17","cited_by_count":32,"area":"alignment"},{"id":"2409.14586","title":"Backtracking Improves Generation Safety","created":"2024-09-22","cited_by_count":32,"area":"alignment"},{"id":"2406.12935","title":"ChatBug: A Common Vulnerability of Aligned LLMs Induced by Chat Templates","created":"2024-06-17","cited_by_count":30,"area":"alignment"},{"id":"2506.08123","title":"QA-LIGN: Aligning LLMs through Constitutionally Decomposed QA","created":"2025-06-09","cited_by_count":28,"area":"alignment"},{"id":"2408.15313","title":"Bi-Factorial Preference Optimization: Balancing Safety-Helpfulness in Language Models","created":"2024-08-27","cited_by_count":27,"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":"2507.04250","title":"Just Enough Shifts: Mitigating Over-Refusal in Aligned Language Models with Targeted Representation Fine-Tuning","created":"2025-07-06","cited_by_count":24,"area":"alignment"},{"id":"2407.08770","title":"Model Surgery: Modulating LLM's Behavior Via Simple Parameter Editing","created":"2024-07-11","cited_by_count":23,"area":"alignment"},{"id":"2407.07342","title":"Multilingual Blending: LLM Safety Alignment Evaluation with Language Mixture","created":"2024-07-10","cited_by_count":22,"area":"alignment"},{"id":"2502.11555","title":"Equilibrate RLHF: Towards Balancing Helpfulness-Safety Trade-off in Large Language Models","created":"2025-02-17","cited_by_count":22,"area":"alignment"},{"id":"2509.11864","title":"NeuroStrike: Neuron-Level Attacks on Aligned LLMs","created":"2025-09-15","cited_by_count":22,"area":"alignment"},{"id":"2501.14940","title":"CASE-Bench: Context-Aware SafEty Benchmark for Large Language Models","created":"2025-01-24","cited_by_count":20,"area":"alignment"},{"id":"2503.18991","title":"Inverse Reinforcement Learning with Dynamic Reward Scaling for LLM Alignment","created":"2025-03-23","cited_by_count":20,"area":"alignment"},{"id":"2506.16447","title":"Probe before You Talk: Towards Black-box Defense against Backdoor Unalignment for Large Language Models","created":"2025-06-19","cited_by_count":19,"area":"alignment"},{"id":"2406.10630","title":"Emerging Safety Attack and Defense in Federated Instruction Tuning of Large Language Models","created":"2024-06-15","cited_by_count":18,"area":"alignment"},{"id":"2503.02574","title":"LLM-Safety Evaluations Lack Robustness","created":"2025-03-04","cited_by_count":18,"area":"alignment"},{"id":"2406.10288","title":"Do as I do (Safely): Mitigating Task-Specific Fine-tuning Risks in Large Language Models","created":"2024-06-12","cited_by_count":17,"area":"alignment"},{"id":"2410.18469","title":"Iterative Self-Tuning LLMs for Enhanced Jailbreaking Capabilities","created":"2024-10-24","cited_by_count":17,"area":"alignment"},{"id":"2502.04322","title":"Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions","created":"2025-02-06","cited_by_count":15,"area":"alignment"},{"id":"2410.10862","title":"Superficial Safety Alignment Hypothesis","created":"2024-10-07","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":"2406.10977","title":"Toward Optimal LLM Alignments Using Two-Player Games","created":"2024-06-16","cited_by_count":13,"area":"alignment"},{"id":"2412.16974","title":"Cannot or Should Not? Automatic Analysis of Refusal Composition in IFT/RLHF Datasets and Refusal Behavior of Black-Box LLMs","created":"2024-12-22","cited_by_count":11,"area":"alignment"},{"id":"2501.16727","title":"xJailbreak: Representation Space Guided Reinforcement Learning for Interpretable LLM Jailbreaking","created":"2025-01-28","cited_by_count":11,"area":"alignment"},{"id":"2508.07172","title":"Gradient Surgery for Safe LLM Fine-Tuning","created":"2025-08-10","cited_by_count":11,"area":"alignment"},{"id":"2408.14866","title":"Advancing Adversarial Suffix Transfer Learning on Aligned Large Language Models","created":"2024-08-27","cited_by_count":10,"area":"alignment"},{"id":"2502.15435","title":"Single-pass Detection of Jailbreaking Input in Large Language Models","created":"2025-02-21","cited_by_count":10,"area":"alignment"},{"id":"2511.03369","title":"Silenced Biases: The Dark Side LLMs Learned to Refuse","created":"2025-11-05","cited_by_count":10,"area":"alignment"},{"id":"2406.18118","title":"SafeAligner: Safety Alignment against Jailbreak Attacks via Response Disparity Guidance","created":"2024-06-26","cited_by_count":9,"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":"2403.14238","title":"Reinforcement Learning from Reflective Feedback (RLRF): Aligning and Improving LLMs via Fine-Grained Self-Reflection","created":"2024-03-21","cited_by_count":8,"area":"alignment"},{"id":"2406.16743","title":"Safety Alignment of Large Language Models via Contrasting Safe and Harmful Distributions","created":"2024-06-24","cited_by_count":8,"area":"alignment"},{"id":"2507.11500","title":"ARMOR: Aligning Secure and Safe Large Language Models via Meticulous Reasoning","created":"2025-07-14","cited_by_count":8,"area":"alignment"},{"id":"2602.11180","title":"Mechanistic Interpretability for Large Language Model Alignment: Progress, Challenges, and Future Directions","created":"2026-01-21","cited_by_count":8,"area":"alignment"},{"id":"2602.06258","title":"GRP-Obliteration: Unaligning LLMs With a Single Unlabeled Prompt","created":"2026-02-05","cited_by_count":7,"area":"alignment"},{"id":"2604.08524","title":"What Drives Representation Steering? A Mechanistic Case Study on Steering Refusal","created":"2026-04-09","cited_by_count":7,"area":"alignment"},{"id":"2504.21053","title":"NeuRel-Attack: Neuron Relearning for Safety Disalignment in Large Language Models","created":"2025-04-29","cited_by_count":6,"area":"alignment"},{"id":"2404.16020","title":"Investigating Adversarial Trigger Transfer in Large Language Models","created":"2024-04-24","cited_by_count":5,"area":"alignment"},{"id":"2411.02785","title":"Stochastic Monkeys at Play: Random Augmentations Cheaply Break LLM Safety Alignment","created":"2024-11-05","cited_by_count":5,"area":"alignment"},{"id":"2406.08842","title":"ContraSolver: Self-Alignment of Language Models by Resolving Internal Preference Contradictions","created":"2024-06-13","cited_by_count":4,"area":"alignment"},{"id":"2410.00451","title":"Unleashing the Unseen: Harnessing Benign Datasets for Jailbreaking Large Language Models","created":"2024-10-01","cited_by_count":4,"area":"alignment"},{"id":"2502.13603","title":"Efficient Safety Retrofitting Against Jailbreaking for LLMs","created":"2025-02-19","cited_by_count":4,"area":"alignment"},{"id":"2510.13698","title":"Attention Misses Visual Risk: Risk-Adaptive Steering for Multimodal Safety Alignment","created":"2025-10-15","cited_by_count":4,"area":"alignment"},{"id":"2601.17642","title":"Health-ORSC-Bench: A Benchmark for Measuring Over-Refusal and Safety Completion in Health Context","created":"2026-01-25","cited_by_count":4,"area":"alignment"},{"id":"2603.15684","title":"State-Dependent Safety Failures in Multi-Turn Language Model Interaction","created":"2026-03-15","cited_by_count":4,"area":"alignment"},{"id":"2605.00195","title":"Diversity in Large Language Models under Supervised Fine-Tuning","created":"2026-04-30","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":"2505.11790","title":"JULI: Jailbreak Large Language Models by Self-Introspection","created":"2025-05-17","cited_by_count":3,"area":"alignment"},{"id":"2506.07402","title":"Beyond Jailbreaks: Revealing Stealthier and Broader LLM Security Risks Stemming from Alignment Failures","created":"2025-06-09","cited_by_count":3,"area":"alignment"},{"id":"2601.19231","title":"LLMs Can Unlearn Refusal with Only 1,000 Benign Samples","created":"2026-01-27","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":"2408.07663","title":"Alignment-Enhanced Decoding:Defending via Token-Level Adaptive Refining of Probability Distributions","created":"2024-08-14","cited_by_count":2,"area":"alignment"},{"id":"2508.14853","title":"Universal and Transferable Adversarial Attack on Large Language Models Using Exponentiated Gradient Descent","created":"2025-08-20","cited_by_count":2,"area":"alignment"},{"id":"2603.05773","title":"Knowing without Acting: The Disentangled Geometry of Safety Mechanisms in Large Language Models","created":"2026-03-06","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":"2606.01196","title":"Low-Resource Safety Failures Are Action Failures, Not Representation Failures","created":"2026-05-31","cited_by_count":2,"area":"alignment"},{"id":"2410.11533","title":"Multi-round jailbreak attack on large language models","created":"2024-10-15","cited_by_count":1,"area":"alignment"},{"id":"2512.06515","title":"ProSocialAlign: Preference Conditioned Test Time Alignment in Language Models","created":"2025-12-06","cited_by_count":1,"area":"alignment"},{"id":"2603.04355","title":"Efficient Refusal Ablation in LLM through Optimal Transport","created":"2026-03-04","cited_by_count":1,"area":"alignment"},{"id":"2605.01899","title":"Disentangling Intent from Role: Adversarial Self-Play for Persona-Invariant Safety Alignment","created":"2026-05-03","cited_by_count":1,"area":"alignment"},{"id":"2605.21706","title":"Latent-space Attacks for Refusal Evasion in Language Models","created":"2026-05-20","cited_by_count":1,"area":"alignment"},{"id":"2605.24154","title":"Palette: A Modular, Controllable, and Efficient Framework for On-demand Authorized Safety Alignment Relaxation in LLMs","created":"2026-05-22","cited_by_count":1,"area":"alignment"},{"id":"2606.28153","title":"Robust Harmful Features Under Jailbreak Attacks: Mechanistic Evidence from Attention Head Specialization in Large Language Models","created":"2026-06-26","cited_by_count":1,"area":"alignment"},{"id":"2502.11379","title":"CCJA: Context-Coherent Jailbreak Attack for Aligned Large Language Models","created":"2025-02-17","cited_by_count":0,"area":"alignment"},{"id":"2506.08885","title":"AdversariaL attacK sAfety aLIgnment(ALKALI): Safeguarding LLMs through GRACE: Geometric Representation-Aware Contrastive Enhancement- Introducing Adversarial Vulnerability Quality Index (AVQI)","created":"2025-06-10","cited_by_count":0,"area":"alignment"},{"id":"2512.05518","title":"Matching Ranks Over Probability Yields Truly Deep Safety Alignment","created":"2025-12-05","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":"2601.06157","title":"ECLIPTICA -- A Framework for Switchable LLM Alignment via CITA - Contrastive Instruction-Tuned Alignment","created":"2026-01-06","cited_by_count":0,"area":"alignment"},{"id":"2602.00619","title":"Jailbreaking LLMs via Calibration","created":"2026-01-31","cited_by_count":0,"area":"alignment"},{"id":"2602.04294","title":"How Few-shot Demonstrations Affect Prompt-based Defenses Against LLM Jailbreak Attacks","created":"2026-02-04","cited_by_count":0,"area":"alignment"},{"id":"2602.08377","title":"Reinforcement Learning with Backtracking Feedback","created":"2026-02-09","cited_by_count":0,"area":"alignment"},{"id":"2604.02574","title":"Understanding the Effects of Safety Unalignment on Large Language Models","created":"2026-04-02","cited_by_count":0,"area":"alignment"},{"id":"2605.08878","title":"Why Do Aligned LLMs Remain Jailbreakable: Refusal-Escape Directions, Operator-Level Sources, and Safety-Utility Trade-off","created":"2026-05-09","cited_by_count":0,"area":"alignment"},{"id":"2607.14147","title":"Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak","created":"2026-07-14","cited_by_count":0,"area":"alignment"},{"id":"2608.10327","title":"Toward a Theory of Value in AI Alignment","created":"2026-08-10","cited_by_count":0,"area":"alignment"},{"id":"2608.11705","title":"Making Your LLMs More Objective: Stabilizing LLM Safety Behavior Across Traits with Trait-Invariant Safety Tuning","created":"2026-08-12","cited_by_count":0,"area":"alignment"},{"id":"2608.17202","title":"Fool's Gold: Defensive Deception Against Safety-Removal Attacks on Open-Weight Models","created":"2026-08-17","cited_by_count":0,"area":"alignment"},{"id":"2608.30197","title":"ALTSTEER: Selective Safety Steering for Moving Beyond Hard Refusals to Constructive Alternatives","created":"2026-08-31","cited_by_count":0,"area":"alignment"},{"id":"2609.02215","title":"ASCII Attack: Recontextualising Harmful Requests as Artistic Critique in Large Language Models","created":"2026-09-02","cited_by_count":0,"area":"alignment"},{"id":"2609.04714","title":"Refuse without Refusal: A Structural Analysis of Safety-Tuning Responses for Reducing False Refusals in Language Models","created":"2026-09-04","cited_by_count":0,"area":"alignment"},{"id":"2609.08634","title":"Suan: Rectifying Direct Preference Safety Alignment in Large Language Models","created":"2026-09-08","cited_by_count":0,"area":"alignment"},{"id":"2609.09793","title":"How Fragile Is Safety Alignment at Frontier Scale? A Single-Direction Attack on a 320B MoE","created":"2026-09-09","cited_by_count":0,"area":"alignment"}],"next_cursor":null}