{"total":35,"rows":[{"id":"2310.10844","title":"Survey of Vulnerabilities in Large Language Models Revealed by Adversarial Attacks","created":"2023-10-16","cited_by_count":194,"area":"safety"},{"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":"2405.15589","title":"Efficient Adversarial Training in LLMs with Continuous Attacks","created":"2024-05-24","cited_by_count":113,"area":"safety"},{"id":"2403.05030","title":"Defending Against Unforeseen Failure Modes with Latent Adversarial Training","created":"2024-03-08","cited_by_count":88,"area":"safety"},{"id":"2505.18333","title":"A Critical Evaluation of Defenses against Prompt Injection Attacks","created":"2025-05-23","cited_by_count":36,"area":"safety"},{"id":"2106.01810","title":"Defending Against Backdoor Attacks in Natural Language Generation","created":"2021-06-03","cited_by_count":33,"area":"safety"},{"id":"2410.22284","title":"Embedding-based classifiers can detect prompt injection attacks","created":"2024-10-29","cited_by_count":27,"area":"safety"},{"id":"2309.13322","title":"From Text to Source: Results in Detecting Large Language Model-Generated Content","created":"2023-09-23","cited_by_count":26,"area":"safety"},{"id":"2504.11168","title":"Bypassing LLM Guardrails: An Empirical Analysis of Evasion Attacks against Prompt Injection and Jailbreak Detection Systems","created":"2025-04-15","cited_by_count":26,"area":"safety"},{"id":"2210.07907","title":"Expose Backdoors on the Way: A Feature-Based Efficient Defense against Textual Backdoor Attacks","created":"2022-10-14","cited_by_count":19,"area":"safety"},{"id":"2409.14740","title":"ToxiCraft: A Novel Framework for Synthetic Generation of Harmful Information","created":"2024-09-23","cited_by_count":17,"area":"safety"},{"id":"2502.02787","title":"SimMark: A Robust Sentence-Level Similarity-Based Watermarking Algorithm for Large Language Models","created":"2025-02-05","cited_by_count":15,"area":"safety"},{"id":"2308.13904","title":"LMSanitator: Defending Prompt-Tuning Against Task-Agnostic Backdoors","created":"2023-08-26","cited_by_count":9,"area":"safety"},{"id":"2502.15427","title":"Adversarial Prompt Evaluation: Systematic Benchmarking of Guardrails Against Prompt Input Attacks on LLMs","created":"2025-02-21","cited_by_count":9,"area":"safety"},{"id":"2510.09694","title":"Kelp: A Streaming Safeguard for Large Models via Latent Dynamics-Guided Risk Detection","created":"2025-10-09","cited_by_count":9,"area":"safety"},{"id":"2305.09574","title":"UOR: Universal Backdoor Attacks on Pre-trained Language Models","created":"2023-05-16","cited_by_count":8,"area":"safety"},{"id":"2506.05446","title":"Sentinel: SOTA model to protect against prompt injections","created":"2025-06-05","cited_by_count":6,"area":"safety"},{"id":"2506.07452","title":"When Style Breaks Safety: Defending LLMs Against Superficial Style Alignment","created":"2025-06-09","cited_by_count":6,"area":"safety"},{"id":"2402.18945","title":"SynGhost: Invisible and Universal Task-agnostic Backdoor Attack via Syntactic Transfer","created":"2024-02-29","cited_by_count":5,"area":"safety"},{"id":"2407.10106","title":"DistillSeq: A Framework for Safety Alignment Testing in Large Language Models using Knowledge Distillation","created":"2024-07-14","cited_by_count":5,"area":"safety"},{"id":"2512.06899","title":"Patronus: Identifying and Mitigating Transferable Backdoors in Pre-trained Language Models","created":"2025-12-07","cited_by_count":4,"area":"safety"},{"id":"2409.19173","title":"HM3: Heterogeneous Multi-Class Model Merging","created":"2024-09-27","cited_by_count":3,"area":"safety"},{"id":"2505.23015","title":"Detecting Stealthy Backdoor Samples based on Intra-class Distance for Large Language Models","created":"2025-05-29","cited_by_count":3,"area":"safety"},{"id":"2605.23196","title":"Prompt Overflow: What the Guardrail Inspects Is Not What the Model Infers","created":"2026-05-22","cited_by_count":3,"area":"safety"},{"id":"2503.05721","title":"What Are They Filtering Out? An Experimental Benchmark of Filtering Strategies for Harm Reduction in Pretraining Datasets","created":"2025-02-17","cited_by_count":2,"area":"safety"},{"id":"2506.05594","title":"SoK: Are Watermarks in LLMs Ready for Deployment?","created":"2025-06-05","cited_by_count":2,"area":"safety"},{"id":"2509.23019","title":"LLM Watermark Evasion via Bias Inversion","created":"2025-09-27","cited_by_count":1,"area":"safety"},{"id":"2602.18583","title":"Luna-2: Scalable Single-Token Evaluation with Small Language Models","created":"2026-02-20","cited_by_count":1,"area":"safety"},{"id":"2603.25176","title":"Prompt Attack Detection with LLM-as-a-Judge and Mixture-of-Models","created":"2026-03-26","cited_by_count":1,"area":"safety"},{"id":"2604.17053","title":"Jailbreaking Large Language Models with Morality Attacks","created":"2026-04-18","cited_by_count":1,"area":"safety"},{"id":"2604.27861","title":"TwinGate: Stateful Defense against Decompositional Jailbreaks in Untraceable Traffic via Asymmetric Contrastive Learning","created":"2026-04-30","cited_by_count":1,"area":"safety"},{"id":"2605.23175","title":"Robust LLM Watermarking with Minimal Semantic Distortion for IP Protection","created":"2026-05-22","cited_by_count":1,"area":"safety"},{"id":"2605.11612","title":"When Emotion Becomes Trigger: Emotion-style dynamic Backdoor Attack Parasitising Large Language Models","created":"2026-05-12","cited_by_count":0,"area":"safety"},{"id":"2605.29659","title":"Opir: Efficient Multi-Task Safety Classification for Toxicity, Jailbreaks, Hate Speech, and Harmful Content","created":"2026-05-28","cited_by_count":0,"area":"safety"},{"id":"2609.15013","title":"Overflip: Repetition-Induced Label Flips in Guardrail Models","created":"2026-09-14","cited_by_count":0,"area":"safety"}],"next_cursor":null}