{"total":112,"rows":[{"id":"2406.11717","title":"Refusal in Language Models Is Mediated by a Single Direction","created":"2024-06-17","cited_by_count":1032,"area":"safety"},{"id":"2305.18462","title":"Membership Inference Attacks against Language Models via Neighbourhood Comparison","created":"2023-05-29","cited_by_count":258,"area":"safety"},{"id":"2108.11577","title":"Machine Unlearning of Features and Labels","created":"2021-08-26","cited_by_count":229,"area":"safety"},{"id":"2303.11570","title":"Boundary Unlearning","created":"2023-03-21","cited_by_count":134,"area":"safety"},{"id":"2208.12268","title":"FedPrompt: Communication-Efficient and Privacy Preserving Prompt Tuning in Federated Learning","created":"2022-08-25","cited_by_count":91,"area":"safety"},{"id":"2310.18491","title":"Publicly-Detectable Watermarking for Language Models","created":"2023-10-27","cited_by_count":76,"area":"safety"},{"id":"2303.03323","title":"CleanCLIP: Mitigating Data Poisoning Attacks in Multimodal Contrastive Learning","created":"2023-03-06","cited_by_count":63,"area":"safety"},{"id":"2405.18540","title":"Learning diverse attacks on large language models for robust red-teaming and safety tuning","created":"2024-05-28","cited_by_count":54,"area":"safety"},{"id":"2502.16580","title":"Can Indirect Prompt Injection Attacks Be Detected and Removed?","created":"2025-02-23","cited_by_count":52,"area":"safety"},{"id":"2406.09073","title":"Are we making progress in unlearning? Findings from the first NeurIPS unlearning competition","created":"2024-06-13","cited_by_count":50,"area":"safety"},{"id":"2401.10371","title":"Langevin Unlearning: A New Perspective of Noisy Gradient Descent for Machine Unlearning","created":"2024-01-18","cited_by_count":49,"area":"safety"},{"id":"2402.00351","title":"Machine Unlearning for Image-to-Image Generative Models","created":"2024-02-01","cited_by_count":47,"area":"safety"},{"id":"2305.06321","title":"SepMark: Deep Separable Watermarking for Unified Source Tracing and Deepfake Detection","created":"2023-05-10","cited_by_count":44,"area":"safety"},{"id":"2212.10717","title":"Hidden Poison: Machine Unlearning Enables Camouflaged Poisoning Attacks","created":"2022-12-21","cited_by_count":43,"area":"safety"},{"id":"2106.15093","title":"Certifiable Machine Unlearning for Linear Models","created":"2021-06-29","cited_by_count":42,"area":"safety"},{"id":"2411.00459","title":"Defense Against Prompt Injection Attack by Leveraging Attack Techniques","created":"2024-11-01","cited_by_count":40,"area":"safety"},{"id":"2208.10836","title":"Evaluating Machine Unlearning via Epistemic Uncertainty","created":"2022-08-23","cited_by_count":39,"area":"safety"},{"id":"2502.18943","title":"Towards Label-Only Membership Inference Attack against Pre-trained Large Language Models","created":"2025-02-26","cited_by_count":37,"area":"safety"},{"id":"2408.00920","title":"Towards Certified Unlearning for Deep Neural Networks","created":"2024-08-01","cited_by_count":31,"area":"safety"},{"id":"2406.16257","title":"Towards Scalable Exact Machine Unlearning Using Parameter-Efficient Fine-Tuning","created":"2024-06-24","cited_by_count":27,"area":"safety"},{"id":"2204.09092","title":"Indiscriminate Data Poisoning Attacks on Neural Networks","created":"2022-04-19","cited_by_count":25,"area":"safety"},{"id":"2303.03592","title":"Exploring the Limits of Model-Targeted Indiscriminate Data Poisoning Attacks","created":"2023-03-07","cited_by_count":23,"area":"safety"},{"id":"2408.00929","title":"Verification of Machine Unlearning is Fragile","created":"2024-08-01","cited_by_count":21,"area":"safety"},{"id":"2110.13541","title":"Qu-ANTI-zation: Exploiting Quantization Artifacts for Achieving Adversarial Outcomes","created":"2021-10-26","cited_by_count":19,"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":"2410.16531","title":"Bayesian scaling laws for in-context learning","created":"2024-10-21","cited_by_count":19,"area":"safety"},{"id":"2506.07596","title":"TwinBreak: Jailbreaking LLM Security Alignments based on Twin Prompts","created":"2025-06-09","cited_by_count":19,"area":"safety"},{"id":"2401.15241","title":"Unlearning Traces the Influential Training Data of Language Models","created":"2024-01-26","cited_by_count":18,"area":"safety"},{"id":"2209.08615","title":"Membership Inference Attacks and Generalization: A Causal Perspective","created":"2022-09-18","cited_by_count":17,"area":"safety"},{"id":"2210.06516","title":"How to Sift Out a Clean Data Subset in the Presence of Data Poisoning?","created":"2022-10-12","cited_by_count":17,"area":"safety"},{"id":"2302.12448","title":"Subspace based Federated Unlearning","created":"2023-02-24","cited_by_count":17,"area":"safety"},{"id":"2406.14796","title":"MU-Bench: A Multitask Multimodal Benchmark for Machine Unlearning","created":"2024-06-21","cited_by_count":17,"area":"safety"},{"id":"2507.13686","title":"TopicAttack: An Indirect Prompt Injection Attack via Topic Transition","created":"2025-07-18","cited_by_count":17,"area":"safety"},{"id":"2504.20472","title":"Robustness via Referencing: Defending against Prompt Injection Attacks by Referencing the Executed Instruction","created":"2025-04-29","cited_by_count":16,"area":"safety"},{"id":"2401.10458","title":"Contrastive Unlearning: A Contrastive Approach to Machine Unlearning","created":"2024-01-19","cited_by_count":15,"area":"safety"},{"id":"2403.17105","title":"Certified Machine Unlearning via Noisy Stochastic Gradient Descent","created":"2024-03-25","cited_by_count":14,"area":"safety"},{"id":"2408.15488","title":"Legilimens: Practical and Unified Content Moderation for Large Language Model Services","created":"2024-08-28","cited_by_count":13,"area":"safety"},{"id":"2208.03309","title":"Lethal Dose Conjecture on Data Poisoning","created":"2022-08-05","cited_by_count":12,"area":"safety"},{"id":"2408.05131","title":"Range Membership Inference Attacks","created":"2024-08-09","cited_by_count":12,"area":"safety"},{"id":"2412.20392","title":"Defending Multimodal Backdoored Models by Repulsive Visual Prompt Tuning","created":"2024-12-29","cited_by_count":12,"area":"safety"},{"id":"2312.16823","title":"Layer Attack Unlearning: Fast and Accurate Machine Unlearning via Layer Level Attack and Knowledge Distillation","created":"2023-12-28","cited_by_count":11,"area":"safety"},{"id":"2508.13650","title":"CRISP: Persistent Concept Unlearning via Sparse Autoencoders","created":"2025-08-19","cited_by_count":11,"area":"safety"},{"id":"2111.09076","title":"To Trust or Not To Trust Prediction Scores for Membership Inference Attacks","created":"2021-11-17","cited_by_count":10,"area":"safety"},{"id":"2302.03684","title":"Temporal Robustness against Data Poisoning","created":"2023-02-07","cited_by_count":10,"area":"safety"},{"id":"2406.12329","title":"Opt-Out: Investigating Entity-Level Unlearning for Large Language Models via Optimal Transport","created":"2024-06-18","cited_by_count":10,"area":"safety"},{"id":"2406.14091","title":"Protecting Privacy Through Approximating Optimal Parameters for Sequence Unlearning in Language Models","created":"2024-06-20","cited_by_count":10,"area":"safety"},{"id":"2407.17710","title":"Revisiting Machine Unlearning with Dimensional Alignment","created":"2024-07-25","cited_by_count":10,"area":"safety"},{"id":"2603.24543","title":"Analysing the Safety Pitfalls of Steering Vectors","created":"2026-03-25","cited_by_count":10,"area":"safety"},{"id":"2412.02454","title":"Gracefully Filtering Backdoor Samples for Generative Large Language Models without Retraining","created":"2024-12-03","cited_by_count":9,"area":"safety"},{"id":"2506.13009","title":"Rectifying Privacy and Efficacy Measurements in Machine Unlearning: A New Inference Attack Perspective","created":"2025-06-16","cited_by_count":9,"area":"safety"},{"id":"2402.12026","title":"Acquiring Clean Language Models from Backdoor Poisoned Datasets by Downscaling Frequency Space","created":"2024-02-19","cited_by_count":8,"area":"safety"},{"id":"2408.09093","title":"BaThe: Defense against the Jailbreak Attack in Multimodal Large Language Models by Treating Harmful Instruction as Backdoor Trigger","created":"2024-08-17","cited_by_count":8,"area":"safety"},{"id":"2410.13274","title":"Breaking Chains: Unraveling the Links in Multi-Hop Knowledge Unlearning","created":"2024-10-17","cited_by_count":8,"area":"safety"},{"id":"2502.19567","title":"Atlas: A Framework for ML Lifecycle Provenance & Transparency","created":"2025-02-26","cited_by_count":8,"area":"safety"},{"id":"2302.14500","title":"FreeEagle: Detecting Complex Neural Trojans in Data-Free Cases","created":"2023-02-28","cited_by_count":7,"area":"safety"},{"id":"2503.02312","title":"Go Beyond Your Means: Unlearning with Per-Sample Gradient Orthogonalization","created":"2025-03-04","cited_by_count":7,"area":"safety"},{"id":"2601.13578","title":"FG-OrIU: Towards Better Forgetting via Feature-Gradient Orthogonality for Incremental Unlearning","created":"2026-01-20","cited_by_count":6,"area":"safety"},{"id":"2405.10425","title":"Data Selection for Transfer Unlearning","created":"2024-05-16","cited_by_count":5,"area":"safety"},{"id":"2405.10612","title":"Not All Prompts Are Secure: A Switchable Backdoor Attack Against Pre-trained Vision Transformers","created":"2024-05-17","cited_by_count":5,"area":"safety"},{"id":"2407.05213","title":"BadCLM: Backdoor Attack in Clinical Language Models for Electronic Health Records","created":"2024-07-06","cited_by_count":5,"area":"safety"},{"id":"2404.04714","title":"Data Poisoning Attacks on Off-Policy Policy Evaluation Methods","created":"2024-04-06","cited_by_count":4,"area":"safety"},{"id":"2503.22413","title":"Instance-Level Data-Use Auditing of Visual ML Models","created":"2025-03-28","cited_by_count":4,"area":"safety"},{"id":"2509.16391","title":"CoUn: Empowering Machine Unlearning via Contrastive Learning","created":"2025-09-19","cited_by_count":4,"area":"safety"},{"id":"2601.22359","title":"The Unseen Threat: Residual Knowledge in Machine Unlearning under Perturbed Samples","created":"2026-01-29","cited_by_count":4,"area":"safety"},{"id":"2602.10568","title":"Gauss-Newton Unlearning for the LLM Era","created":"2026-02-11","cited_by_count":4,"area":"safety"},{"id":"2206.05565","title":"NeuGuard: Lightweight Neuron-Guided Defense against Membership Inference Attacks","created":"2022-06-11","cited_by_count":3,"area":"safety"},{"id":"2402.10065","title":"Some Targets Are Harder to Identify than Others: Quantifying the Target-dependent Membership Leakage","created":"2024-02-15","cited_by_count":3,"area":"safety"},{"id":"2501.16534","title":"Targeting Alignment: Extracting Safety Classifiers of Aligned LLMs","created":"2025-01-27","cited_by_count":3,"area":"safety"},{"id":"2502.12464","title":"SafeRoute: Adaptive Model Selection for Efficient and Accurate Safety Guardrails in Large Language Models","created":"2025-02-18","cited_by_count":3,"area":"safety"},{"id":"2503.09330","title":"Group-robust Machine Unlearning","created":"2025-03-12","cited_by_count":3,"area":"safety"},{"id":"2508.17247","title":"Uncovering and Mitigating Destructive Multi-Embedding Attacks in Deepfake Proactive Forensics","created":"2025-08-24","cited_by_count":3,"area":"safety"},{"id":"2510.03705","title":"Backdoor-Powered Prompt Injection Attacks Nullify Defense Methods","created":"2025-10-04","cited_by_count":3,"area":"safety"},{"id":"2308.09971","title":"Disposable Transfer Learning for Selective Source Task Unlearning","created":"2023-08-19","cited_by_count":2,"area":"safety"},{"id":"2404.18567","title":"Double Backdoored: Converting Code Large Language Model Backdoors to Traditional Malware via Adversarial Instruction Tuning Attacks","created":"2024-04-29","cited_by_count":2,"area":"safety"},{"id":"2407.07539","title":"Machine Unlearning for Medical Imaging","created":"2024-07-10","cited_by_count":2,"area":"safety"},{"id":"2407.13796","title":"Continuous Embedding Attacks via Clipped Inputs in Jailbreaking Large Language Models","created":"2024-07-16","cited_by_count":2,"area":"safety"},{"id":"2502.06374","title":"Hyperparameters in Score-Based Membership Inference Attacks","created":"2025-02-10","cited_by_count":2,"area":"safety"},{"id":"2505.08255","title":"Where the Devil Hides: Deepfake Detectors Can No Longer Be Trusted","created":"2025-05-13","cited_by_count":2,"area":"safety"},{"id":"2505.15175","title":"A Linear Approach to Data Poisoning","created":"2025-05-21","cited_by_count":2,"area":"safety"},{"id":"2603.02588","title":"ExpGuard: LLM Content Moderation in Specialized Domains","created":"2026-03-03","cited_by_count":2,"area":"safety"},{"id":"2604.08297","title":"Towards Identification and Intervention of Safety-Critical Parameters in Large Language Models","created":"2026-04-09","cited_by_count":2,"area":"safety"},{"id":"2109.04566","title":"SanitAIs: Unsupervised Data Augmentation to Sanitize Trojaned Neural Networks","created":"2021-09-09","cited_by_count":1,"area":"safety"},{"id":"2305.11347","title":"Quantifying the robustness of deep multispectral segmentation models against natural perturbations and data poisoning","created":"2023-05-18","cited_by_count":1,"area":"safety"},{"id":"2306.03600","title":"Avoid Adversarial Adaption in Federated Learning by Multi-Metric Investigations","created":"2023-06-06","cited_by_count":1,"area":"safety"},{"id":"2405.15140","title":"Better Membership Inference Privacy Measurement through Discrepancy","created":"2024-05-24","cited_by_count":1,"area":"safety"},{"id":"2408.02813","title":"Mitigating Malicious Attacks in Federated Learning via Confidence-aware Defense","created":"2024-08-05","cited_by_count":1,"area":"safety"},{"id":"2410.08069","title":"Unlearning-based Neural Interpretations","created":"2024-10-10","cited_by_count":1,"area":"safety"},{"id":"2506.09923","title":"Apollo: A Posteriori Label-Only Membership Inference Attack Towards Machine Unlearning","created":"2025-06-11","cited_by_count":1,"area":"safety"},{"id":"2509.15861","title":"ToFU: Transforming How Federated Learning Systems Forget User Data","created":"2025-09-19","cited_by_count":1,"area":"safety"},{"id":"2509.17773","title":"LoT-Pass: Long-term-robust Image Watermarking for Image to Video Generation","created":"2025-09-22","cited_by_count":1,"area":"safety"},{"id":"2510.18019","title":"Is Multilingual LLM Watermarking Truly Multilingual? Scaling Robustness to 100+ Languages via Back-Translation","created":"2025-10-20","cited_by_count":1,"area":"safety"},{"id":"2511.19566","title":"ModHiFi: Identifying High Fidelity predictive components for Model Modification","created":"2025-11-24","cited_by_count":1,"area":"safety"},{"id":"2512.13317","title":"Face Identity Unlearning for Retrieval via Embedding Dispersion","created":"2025-12-15","cited_by_count":1,"area":"safety"},{"id":"2512.16126","title":"Dual-View Inference Attack: Machine Unlearning Amplifies Privacy Exposure","created":"2025-12-18","cited_by_count":1,"area":"safety"},{"id":"2202.10309","title":"HoneyModels: Machine Learning Honeypots","created":"2022-02-21","cited_by_count":0,"area":"safety"},{"id":"2310.16453","title":"ClearMark: Intuitive and Robust Model Watermarking via Transposed Model Training","created":"2023-10-25","cited_by_count":0,"area":"safety"},{"id":"2311.04009","title":"AGNES: Abstraction-guided Framework for Deep Neural Networks Security","created":"2023-11-07","cited_by_count":0,"area":"safety"},{"id":"2402.12189","title":"Amplifying Training Data Exposure through Fine-Tuning with Pseudo-Labeled Memberships","created":"2024-02-19","cited_by_count":0,"area":"safety"},{"id":"2511.18660","title":"Subtract the Corruption: Training-Data-Free Corrective Machine Unlearning using Task Arithmetic","created":"2025-11-24","cited_by_count":0,"area":"safety"},{"id":"2512.05254","title":"When unlearning is free: leveraging low influence points to reduce computational costs","created":"2025-12-04","cited_by_count":0,"area":"safety"}],"next_cursor":"100"}