{"total":80,"rows":[{"id":"2407.15549","title":"Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs","created":"2024-07-22","cited_by_count":173,"area":"safety"},{"id":"2510.05179","title":"Agentic Misalignment: How LLMs Could Be Insider Threats","created":"2025-10-05","cited_by_count":158,"area":"agents"},{"id":"2408.15221","title":"LLM Defenses Are Not Robust to Multi-Turn Human Jailbreaks Yet","created":"2024-08-27","cited_by_count":151,"area":"safety"},{"id":"2504.15585","title":"A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment","created":"2025-04-22","cited_by_count":116,"area":"lr"},{"id":"2409.01586","title":"Booster: Tackling Harmful Fine-tuning for Large Language Models via Attenuating Harmful Perturbation","created":"2024-09-03","cited_by_count":92,"area":null},{"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":"2410.02832","title":"FlipAttack: Jailbreak LLMs via Flipping","created":"2024-10-02","cited_by_count":80,"area":"safety"},{"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":"2505.02077","title":"Open Challenges in Multi-Agent Security: Towards Secure Systems of Interacting AI Agents","created":"2025-05-04","cited_by_count":71,"area":"safety"},{"id":"2407.09121","title":"Refuse Whenever You Feel Unsafe: Improving Safety in LLMs via Decoupled Refusal Training","created":"2024-07-12","cited_by_count":69,"area":null},{"id":"2410.03768","title":"Hidden in Plain Text: Emergence & Mitigation of Steganographic Collusion in LLMs","created":"2024-10-02","cited_by_count":55,"area":null},{"id":"2412.07097","title":"On Evaluating the Durability of Safeguards for Open-Weight LLMs","created":"2024-12-10","cited_by_count":52,"area":null},{"id":"2502.05209","title":"Model Tampering Attacks Enable More Rigorous Evaluations of LLM Capabilities","created":"2025-02-03","cited_by_count":39,"area":null},{"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":"2408.02946","title":"Scaling Trends for Data Poisoning in LLMs","created":"2024-08-06","cited_by_count":30,"area":"safety"},{"id":"2505.06843","title":"Benign Samples Matter! Fine-tuning On Outlier Benign Samples Severely Breaks Safety","created":"2025-05-11","cited_by_count":22,"area":"safety"},{"id":"2501.18100","title":"Panacea: Mitigating Harmful Fine-tuning for Large Language Models via Post-fine-tuning Perturbation","created":"2025-01-30","cited_by_count":20,"area":null},{"id":"2504.05259","title":"How to evaluate control measures for LLM agents? A trajectory from today to superintelligence","created":"2025-04-07","cited_by_count":20,"area":"agents"},{"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":"2410.01294","title":"Endless Jailbreaks with Bijection Learning","created":"2024-10-02","cited_by_count":19,"area":"safety"},{"id":"2506.22521","title":"A Survey on Model Extraction Attacks and Defenses for Large Language Models","created":"2025-06-26","cited_by_count":19,"area":null},{"id":"2508.03153","title":"Estimating Worst-Case Frontier Risks of Open-Weight LLMs","created":"2025-08-05","cited_by_count":19,"area":null},{"id":"2602.04863","title":"Subliminal Effects in Your Data: A General Mechanism via Log-Linearity","created":"2026-02-04","cited_by_count":19,"area":null},{"id":"2501.09798","title":"Fun-tuning: Characterizing the Vulnerability of Proprietary LLMs to Optimization-based Prompt Injection Attacks via the Fine-Tuning Interface","created":"2025-01-16","cited_by_count":17,"area":"safety"},{"id":"2502.08301","title":"Compromising Honesty and Harmlessness in Language Models via Deception Attacks","created":"2025-02-12","cited_by_count":17,"area":null},{"id":"2510.27172","title":"Adaptive Defense against Harmful Fine-Tuning for Large Language Models via Bayesian Data Scheduler","created":"2025-10-31","cited_by_count":12,"area":null},{"id":"2505.16559","title":"CTRAP: Embedding Collapse Trap to Safeguard Large Language Models from Harmful Fine-Tuning","created":"2025-05-22","cited_by_count":11,"area":"safety"},{"id":"2508.17158","title":"Towards Safeguarding LLM Fine-tuning APIs against Cipher Attacks","created":"2025-08-23","cited_by_count":11,"area":null},{"id":"2604.16812","title":"Introspection Adapters: Training LLMs to Report Their Learned Behaviors","created":"2026-04-18","cited_by_count":11,"area":null},{"id":"2505.12186","title":"Self-Destructive Language Model","created":"2025-05-18","cited_by_count":10,"area":"safety"},{"id":"2510.02833","title":"Attack via Overfitting: 10-shot Benign Fine-tuning to Jailbreak LLMs","created":"2025-10-03","cited_by_count":10,"area":"alignment"},{"id":"2512.10150","title":"Unforgotten Safety: Preserving Safety Alignment of Large Language Models with Continual Learning","created":"2025-12-10","cited_by_count":10,"area":"alignment"},{"id":"2604.05432","title":"Your LLM Agent Can Leak Your Data: Data Exfiltration via Backdoored Tool Use","created":"2026-04-07","cited_by_count":10,"area":"agents"},{"id":"2507.11630","title":"Jailbreak-Tuning: Models Efficiently Learn Jailbreak Susceptibility","created":"2025-07-15","cited_by_count":9,"area":"safety"},{"id":"2506.06414","title":"Benchmarking Misuse Mitigation Against Covert Adversaries","created":"2025-06-06","cited_by_count":8,"area":null},{"id":"2510.09714","title":"All Code, No Thought: Current Language Models Struggle to Reason in Ciphered Language","created":"2025-10-10","cited_by_count":8,"area":"lr"},{"id":"2503.19041","title":"LookAhead Tuning: Safer Language Models via Partial Answer Previews","created":"2025-03-24","cited_by_count":7,"area":"alignment"},{"id":"2409.03274","title":"Recent Advances in Attack and Defense Approaches of Large Language Models","created":"2024-09-05","cited_by_count":6,"area":null},{"id":"2502.19537","title":"No, of Course I Can! Deeper Fine-Tuning Attacks That Bypass Token-Level Safety Mechanisms","created":"2025-02-26","cited_by_count":6,"area":"safety"},{"id":"2410.16672","title":"The Tug of War Within: Mitigating the Fairness-Privacy Conflicts in Large Language Models","created":"2024-10-22","cited_by_count":5,"area":null},{"id":"2505.16567","title":"Watch your steps: Dormant Adversarial Behaviors that Activate upon LLM Finetuning","created":"2025-05-22","cited_by_count":5,"area":"safety"},{"id":"2510.04891","title":"SocialHarmBench: Revealing LLM Vulnerabilities to Socially Harmful Requests","created":"2025-10-06","cited_by_count":5,"area":"safety"},{"id":"2602.05228","title":"Surgery: Mitigating Harmful Fine-Tuning for Large Language Models via Attention Sink","created":"2026-02-05","cited_by_count":4,"area":"alignment"},{"id":"2412.03824","title":"Towards Data Governance of Frontier AI Models","created":"2024-12-05","cited_by_count":3,"area":null},{"id":"2503.17388","title":"AI Companies Should Report Pre- and Post-Mitigation Safety Evaluations","created":"2025-03-17","cited_by_count":3,"area":null},{"id":"2505.11790","title":"JULI: Jailbreak Large Language Models by Self-Introspection","created":"2025-05-17","cited_by_count":3,"area":"alignment"},{"id":"2507.21170","title":"OneShield -- the Next Generation of LLM Guardrails","created":"2025-07-25","cited_by_count":3,"area":"safety"},{"id":"2510.16255","title":"Detecting Adversarial Fine-tuning with Auditing Agents","created":"2025-10-17","cited_by_count":3,"area":"agents"},{"id":"2602.16729","title":"Intent Laundering: AI Safety Datasets Are Not What They Seem","created":"2026-02-17","cited_by_count":3,"area":null},{"id":"2604.14865","title":"Segment-Level Coherence for Robust Harmful Intent Probing in LLMs","created":"2026-04-16","cited_by_count":3,"area":"safety"},{"id":"2510.01342","title":"Fine-Tuning Jailbreaks under Highly Constrained Black-Box Settings: A Three-Pronged Approach","created":"2025-10-01","cited_by_count":2,"area":"safety"},{"id":"2510.07626","title":"LLM Unlearning Under the Microscope: A Full-Stack View on Methods and Metrics","created":"2025-10-08","cited_by_count":2,"area":"safety"},{"id":"2601.22710","title":"AlienLM: Alienization of Language for API-Boundary Privacy in Black-Box LLMs","created":"2026-01-30","cited_by_count":2,"area":null},{"id":"2602.06911","title":"TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering","created":"2026-02-06","cited_by_count":2,"area":"safety"},{"id":"2603.08104","title":"Invisible Safety Threat: Malicious Finetuning for LLM via Steganography","created":"2026-03-09","cited_by_count":2,"area":"alignment"},{"id":"2508.20333","title":"Poison Once, Refuse Forever: Weaponizing Alignment for Injecting Bias in LLMs","created":"2025-08-28","cited_by_count":1,"area":null},{"id":"2509.02411","title":"A Survey: Towards Privacy and Security in Mobile Large Language Models","created":"2025-09-02","cited_by_count":1,"area":null},{"id":"2511.02356","title":"ASTRA: An Automated Framework for Strategy Discovery, Retrieval, and Evolution for Jailbreaking LLMs","created":"2025-11-04","cited_by_count":1,"area":"alignment"},{"id":"2601.22818","title":"Hide and Seek in Embedding Space: Geometry-based Steganography and Detection in Large Language Models","created":"2026-01-30","cited_by_count":1,"area":"mechinterp"},{"id":"2602.13033","title":"Buy versus Build an LLM: A Decision Framework for Governments","created":"2026-02-13","cited_by_count":1,"area":null},{"id":"2603.24857","title":"AI Security in the Foundation Model Era: A Comprehensive Survey from a Unified Perspective","created":"2026-03-25","cited_by_count":1,"area":"safety"},{"id":"2603.29038","title":"Trojan-Speak: Bypassing Constitutional Classifiers with No Jailbreak Tax via Adversarial Finetuning","created":"2026-03-30","cited_by_count":1,"area":"safety"},{"id":"2605.25073","title":"Security in the Fine-Tuning Lifecycle of Large Language Models: Threats, Defenses,Evaluation, and Future Directions","created":"2026-05-24","cited_by_count":1,"area":"safety"},{"id":"2606.17110","title":"Loss Landscape Poisoning: Targeted Extraction of Unseen Training Data from LLMs","created":"2026-06-15","cited_by_count":1,"area":null},{"id":"2609.01091","title":"Subliminal Learning as Trait-Direction Drift: A Mechanism and Targeted Control under SFT Distillation","created":"2026-09-01","cited_by_count":1,"area":"distill"},{"id":"2602.17546","title":"Learning to Stay Safe: Adaptive Regularization Against Safety Degradation during Fine-Tuning","created":"2026-02-19","cited_by_count":0,"area":null},{"id":"2604.00324","title":"The Persistent Vulnerability of Aligned AI Systems","created":"2026-03-31","cited_by_count":0,"area":null},{"id":"2604.07754","title":"The Art of (Mis)alignment: How Fine-Tuning Methods Effectively Misalign and Realign LLMs in Post-Training","created":"2026-04-09","cited_by_count":0,"area":"alignment"},{"id":"2605.10998","title":"Few-Shot Truly Benign DPO Attack for Jailbreaking LLMs","created":"2026-05-09","cited_by_count":0,"area":"alignment"},{"id":"2605.14194","title":"GradShield: Alignment Preserving Finetuning","created":"2026-05-13","cited_by_count":0,"area":null},{"id":"2606.09411","title":"Now You (Still) See Me: Detecting Evasive Steganographic Payloads in LLMs","created":"2026-06-08","cited_by_count":0,"area":null},{"id":"2608.30619","title":"Hidden Threat in Synthetic Data: Covert Targeted Bias Injection through Benign Text","created":"2026-08-31","cited_by_count":0,"area":"synth"},{"id":"2609.05241","title":"Uncensored Open-weight Models: Redistribution as the Persistence Layer","created":"2026-09-04","cited_by_count":0,"area":"safety"},{"id":"2609.06851","title":"You Are What You Read: Misalignment via In-Context Persona Induction","created":"2026-09-06","cited_by_count":0,"area":null},{"id":"2609.09553","title":"Arbitrary Cipher Attacks Against Large Language Models Do Not Require Fine-Tuning","created":"2026-09-09","cited_by_count":0,"area":"safety"},{"id":"2609.22144","title":"Multilingual Safety Signals Are Multi-Layered: Filtering Safety-Degrading Data for Safer LLMs","created":"2026-08-26","cited_by_count":0,"area":"alignment"},{"id":"2609.29960","title":"Beyond Average Safety: Chance-Constrained LLM Fine-tuning","created":"2026-09-24","cited_by_count":0,"area":null},{"id":"2610.03033","title":"When Numbers Start Talking: Numerical Signalling and Strategic Behaviour Among LLMs","created":"2026-10-02","cited_by_count":0,"area":null},{"id":"2610.07518","title":"Harmful SFT Leaves a Continuous Trace in LLM Checkpoint Updates","created":"2026-10-05","cited_by_count":0,"area":null},{"id":"2610.12644","title":"Cancerous Backdoor Attack: Gradual Activation via Invisible Tokens and A Stage Controller","created":"2026-10-08","cited_by_count":0,"area":"safety"}],"next_cursor":null}