{"total":97,"rows":[{"id":"2501.09004","title":"Aegis2.0: A Diverse AI Safety Dataset and Risks Taxonomy for Alignment of LLM Guardrails","created":"2025-01-15","cited_by_count":132,"area":"safety"},{"id":"2411.00348","title":"Attention Tracker: Detecting Prompt Injection Attacks in LLMs","created":"2024-11-01","cited_by_count":95,"area":"safety"},{"id":"2506.07022","title":"AlphaSteer: Learning Refusal Steering with Principled Null-Space Constraint","created":"2025-06-08","cited_by_count":49,"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":"2506.09443","title":"LLMs Cannot Reliably Judge (Yet?): A Comprehensive Assessment on the Robustness of LLM-as-a-Judge","created":"2025-06-11","cited_by_count":37,"area":"safety"},{"id":"2412.10321","title":"AdvPrefix: An Objective for Nuanced LLM Jailbreaks","created":"2024-12-13","cited_by_count":25,"area":"safety"},{"id":"2504.07887","title":"Benchmarking Adversarial Robustness to Bias Elicitation in Large Language Models: Scalable Automated Assessment with LLM-as-a-Judge","created":"2025-04-10","cited_by_count":23,"area":"safety"},{"id":"2504.18530","title":"Scaling Laws For Scalable Oversight","created":"2025-04-25","cited_by_count":23,"area":"safety"},{"id":"2502.11190","title":"ReLearn: Unlearning via Learning for Large Language Models","created":"2025-02-16","cited_by_count":22,"area":"safety"},{"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":"2410.02163","title":"Adversarial Decoding: Generating Readable Documents for Adversarial Objectives","created":"2024-10-03","cited_by_count":16,"area":"safety"},{"id":"2412.03235","title":"Does Safety Training of LLMs Generalize to Semantically Related Natural Prompts?","created":"2024-12-04","cited_by_count":15,"area":"safety"},{"id":"2412.16633","title":"Easier Said Than Done: Unpacking Intent-Behavior Gap in Jailbreaking LLM-Based Robots","created":"2024-12-21","cited_by_count":15,"area":"safety"},{"id":"2503.08195","title":"Dialogue Injection Attack: Jailbreaking LLMs through Context Manipulation","created":"2025-03-11","cited_by_count":14,"area":"safety"},{"id":"2512.07761","title":"TROJail: Trajectory-Level Optimization for Multi-Turn Large Language Model Jailbreaks with Process Rewards","created":"2025-12-08","cited_by_count":14,"area":"safety"},{"id":"2504.15047","title":"RainbowPlus: Enhancing Adversarial Prompt Generation via Evolutionary Quality-Diversity Search","created":"2025-04-21","cited_by_count":11,"area":"safety"},{"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":"2506.06278","title":"Distillation Robustifies Unlearning","created":"2025-06-06","cited_by_count":11,"area":"safety"},{"id":"2412.00765","title":"SelfPrompt: Autonomously Evaluating LLM Robustness via Domain-Constrained Knowledge Guidelines and Refined Adversarial Prompts","created":"2024-12-01","cited_by_count":10,"area":"safety"},{"id":"2502.15086","title":"Is Safety Standard Same for Everyone? User-Specific Safety Evaluation of Large Language Models","created":"2025-02-20","cited_by_count":10,"area":"safety"},{"id":"2506.12880","title":"Universal Jailbreak Suffixes Are Strong Attention Hijackers","created":"2025-06-15","cited_by_count":10,"area":"safety"},{"id":"2505.23814","title":"Watermarking Without Standards Is Not AI Governance","created":"2025-05-27","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":"2510.02286","title":"Tree-based Dialogue Reinforced Policy Optimization for Red-Teaming Attacks","created":"2025-10-02","cited_by_count":8,"area":"safety"},{"id":"2609.29429","title":"Just Ask Jev: Reinforcement Learning for Calibrated Decisions as a Zero-Shot Detector of AI Alignment Failures","created":"2026-09-24","cited_by_count":8,"area":"safety"},{"id":"2501.00055","title":"LLM-Virus: Evolutionary Jailbreak Attack on Large Language Models","created":"2024-12-28","cited_by_count":7,"area":"safety"},{"id":"2501.13677","title":"HumorReject: Decoupling LLM Safety from Refusal Prefix via A Little Humor","created":"2025-01-23","cited_by_count":6,"area":"safety"},{"id":"2502.19207","title":"FaithUn: Toward Faithful Forgetting in Language Models by Investigating the Interconnectedness of Knowledge","created":"2025-02-26","cited_by_count":6,"area":"safety"},{"id":"2503.05445","title":"Are Your LLM-based Text-to-SQL Models Secure? Exploring SQL Injection via Backdoor Attacks","created":"2025-03-07","cited_by_count":6,"area":"safety"},{"id":"2505.14368","title":"From ASR to ASP: Evaluating Prompt Attack Vulnerabilities Against Open-Source LLMs","created":"2025-05-20","cited_by_count":6,"area":"safety"},{"id":"2602.12418","title":"Sparse Autoencoders are Capable LLM Jailbreak Mitigators","created":"2026-02-12","cited_by_count":6,"area":"safety"},{"id":"2603.00498","title":"Antibody: Strengthening Defense Against Harmful Fine-Tuning for Large Language Models via Attenuating Harmful Gradient Influence","created":"2026-02-28","cited_by_count":6,"area":"safety"},{"id":"2510.09871","title":"CoBia: Constructed Conversations Can Trigger Otherwise Concealed Societal Biases in LLMs","created":"2025-10-10","cited_by_count":5,"area":"safety"},{"id":"2503.01926","title":"Unnatural Languages Are Not Bugs but Features for LLMs","created":"2025-03-02","cited_by_count":4,"area":"safety"},{"id":"2503.24191","title":"When Grammar Guides the Attack: Uncovering Control-Plane Vulnerabilities in LLMs with Structured Output","created":"2025-03-31","cited_by_count":4,"area":"safety"},{"id":"2507.11112","title":"Multi-Trigger Poisoning Amplifies Backdoor Vulnerabilities in LLMs","created":"2025-07-15","cited_by_count":4,"area":"safety"},{"id":"2508.03829","title":"Majority Bit-Aware Watermarking For Large Language Models","created":"2025-08-05","cited_by_count":4,"area":"safety"},{"id":"2502.10487","title":"Fast Proxies for LLM Robustness Evaluation","created":"2025-02-14","cited_by_count":3,"area":"safety"},{"id":"2509.15631","title":"Sparse-Autoencoder-Guided Internal Representation Unlearning for Large Language Models","created":"2025-09-19","cited_by_count":3,"area":"safety"},{"id":"2601.15111","title":"Auditing Language Model Unlearning via Information Decomposition","created":"2026-01-21","cited_by_count":3,"area":"safety"},{"id":"2602.20170","title":"CAGE: A Framework for Culturally Adaptive Red-Teaming Benchmark Generation","created":"2026-02-09","cited_by_count":3,"area":"safety"},{"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":"2505.01315","title":"Helping Large Language Models Protect Themselves: An Enhanced Filtering and Summarization System","created":"2025-05-02","cited_by_count":2,"area":"safety"},{"id":"2506.07356","title":"Safety-Aligned Weights Are Not Enough: Refusal-Teacher-Guided Finetuning Enhances Safety and Downstream Performance under Harmful Finetuning Attacks","created":"2025-06-09","cited_by_count":2,"area":"safety"},{"id":"2602.18782","title":"MANATEE: Inference-Time Lightweight Diffusion Based Safety Defense for LLMs","created":"2026-02-21","cited_by_count":2,"area":"safety"},{"id":"2605.05443","title":"SLAM: Structural Linguistic Activation Marking for Language Models","created":"2026-05-06","cited_by_count":2,"area":"safety"},{"id":"2606.08044","title":"When Behavioral Safety Evaluation Fails: A Representation-Level Perspective","created":"2026-06-06","cited_by_count":2,"area":"safety"},{"id":"2503.03502","title":"Geometry-Guided Adversarial Prompt Detection via Curvature and Local Intrinsic Dimension","created":"2025-03-05","cited_by_count":1,"area":"safety"},{"id":"2505.11837","title":"On Membership Inference Attacks in Knowledge Distillation","created":"2025-05-17","cited_by_count":1,"area":"safety"},{"id":"2506.15076","title":"Learning-Time Encoding Shapes Unlearning in LLMs","created":"2025-06-18","cited_by_count":1,"area":"safety"},{"id":"2507.00026","title":"RedTopic: Toward Topic-Diverse Red Teaming of Large Language Models","created":"2025-06-17","cited_by_count":1,"area":"safety"},{"id":"2509.15621","title":"Concept Unlearning in Large Language Models via Self-Constructed Knowledge Triplets","created":"2025-09-19","cited_by_count":1,"area":"safety"},{"id":"2509.16462","title":"Let's Unlearn Stereotypes Before Decision-Making: Assessing the Impact of Intrinsic Bias Mitigation on Downstream Fairness in LLMs","created":"2025-09-19","cited_by_count":1,"area":"safety"},{"id":"2509.25843","title":"ASGuard: Activation-Scaling Guard to Mitigate Targeted Jailbreaking Attack","created":"2025-09-30","cited_by_count":1,"area":"safety"},{"id":"2604.18970","title":"Mechanistic Anomaly Detection via Functional Attribution","created":"2026-04-21","cited_by_count":1,"area":"safety"},{"id":"2606.19887","title":"FinRED: An Expert-Guided Benchmark Generation and Evaluation Framework for Financial LLM Red-Teaming","created":"2026-06-18","cited_by_count":1,"area":"safety"},{"id":"2607.27081","title":"On-Policy Distillation for LLM Safety: A Routing Approach to Template-Robust Realignment","created":"2026-07-29","cited_by_count":1,"area":"safety"},{"id":"2608.04415","title":"Social Pressure Breaks Majority Voting in LLM Safety Panels","created":"2026-08-05","cited_by_count":1,"area":"safety"},{"id":"2608.23028","title":"PsychJail: Exploring Psychological Jailbreaks via Multi-Turn Persuasion of LLM Policies","created":"2026-08-24","cited_by_count":1,"area":"safety"},{"id":"2609.16204","title":"Decoy Direction Optimization: A Post-Hoc Defense Against LLM Abliteration","created":"2026-09-14","cited_by_count":1,"area":"safety"},{"id":"2501.18628","title":"TombRaider: Entering the Vault of History to Jailbreak Large Language Models","created":"2025-01-27","cited_by_count":0,"area":"safety"},{"id":"2505.07167","title":"Decoding One Safety Trigger Token for Balancing Safety and Usability in Large Language Models","created":"2025-05-12","cited_by_count":0,"area":"safety"},{"id":"2505.13819","title":"Fragments to Facts: Partial-Information Fragment Inference from LLMs","created":"2025-05-20","cited_by_count":0,"area":"safety"},{"id":"2509.23381","title":"Guard Vector: Beyond English LLM Guardrails with Task-Vector Composition and Streaming-Aware Prefix SFT","created":"2025-09-27","cited_by_count":0,"area":"safety"},{"id":"2510.12672","title":"Keep Calm and Avoid Harmful Content: Concept Alignment and Latent Manipulation Towards Safer Answers","created":"2025-10-14","cited_by_count":0,"area":"safety"},{"id":"2602.01752","title":"WorldCup Sampling for Multi-bit LLM Watermarking","created":"2026-02-02","cited_by_count":0,"area":"safety"},{"id":"2602.08877","title":"Stress-Testing Alignment Audits With Prompt-Level Strategic Deception","created":"2026-02-09","cited_by_count":0,"area":"safety"},{"id":"2604.10403","title":"Latent Instruction Representation Alignment: defending against jailbreaks, backdoors and undesired knowledge in LLMs","created":"2026-04-12","cited_by_count":0,"area":"safety"},{"id":"2604.23130","title":"From Concept-Aligned Tokens to Vulnerable Features: Mechanistic Localization of Jailbreaks","created":"2026-04-25","cited_by_count":0,"area":"safety"},{"id":"2605.24550","title":"Jailbreak to Protect: Buffering and Reinforcing via Temporary Jailbreaking for Safe Fine-Tuning in Large Language Models","created":"2026-05-23","cited_by_count":0,"area":"safety"},{"id":"2605.26409","title":"Jailbreak susceptibility prediction and mitigation via the behavioral geometry of models","created":"2026-05-26","cited_by_count":0,"area":"safety"},{"id":"2605.29629","title":"Beyond Attack Success Rate: Temporal Logit Observability for LLM Safety Failures","created":"2026-05-28","cited_by_count":0,"area":"safety"},{"id":"2606.03486","title":"NeuroArmor: Safe-Variant-Guided Representation Consistency for Selective Re-Anchoring in Jailbreak Defense","created":"2026-06-02","cited_by_count":0,"area":"safety"},{"id":"2606.05614","title":"Safety Paradox: How Enhanced Safety Awareness Leaves LLMs Vulnerable to Posterior Attack","created":"2026-06-04","cited_by_count":0,"area":"safety"},{"id":"2606.05958","title":"Steering Vectors are an Adversarial Attack Surface","created":"2026-06-04","cited_by_count":0,"area":"safety"},{"id":"2606.09388","title":"Distilling Safe LLM Systems via Soft Prompts for On Device Settings","created":"2026-06-08","cited_by_count":0,"area":"safety"},{"id":"2606.24267","title":"Pigeonholing: how bad prompts hurt models, causing collapse and mistakes","created":"2026-06-23","cited_by_count":0,"area":"safety"},{"id":"2607.11070","title":"MJ: Multi-turn LLM Jailbreaking via Decomposed Credit Assignment","created":"2026-07-13","cited_by_count":0,"area":"safety"},{"id":"2607.13088","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","created":"2026-07-13","cited_by_count":0,"area":"safety"},{"id":"2608.10621","title":"ProbGuard: Calibrated Safety Risk Estimation from LLM Output Distributions","created":"2026-08-11","cited_by_count":0,"area":"safety"},{"id":"2608.22218","title":"PURA: Provably Unbiased and Robust Multi-Bit Watermarking for AI-Generated Text Attribution","created":"2026-08-23","cited_by_count":0,"area":"safety"},{"id":"2608.30403","title":"Why Are LLM Backdoor Defenses Fragmented? A Feature-Level Explanation with Sparse Autoencoders","created":"2026-08-31","cited_by_count":0,"area":"safety"},{"id":"2608.30585","title":"The Safety Relay in Roleplay Jailbreaks: A Component-Resolved Causal Analysis of Harm Recognition and Refusal","created":"2026-08-31","cited_by_count":0,"area":"safety"},{"id":"2609.02177","title":"WeaveMark: Robust and Scalable Multi-bit LLM Watermarking via Coded Payload Spreading","created":"2026-09-02","cited_by_count":0,"area":"safety"},{"id":"2609.05074","title":"Influence Score and Transformers interpretability: Measure of the Effective Impact of Attention Heads at inference time","created":"2026-09-04","cited_by_count":0,"area":"safety"},{"id":"2609.06540","title":"SRD-GUARD: A Defense Framework of LLMs via Semantic Rewriting and Joint Multi-Model Scoring for Latent Intent Exposure","created":"2026-09-06","cited_by_count":0,"area":"safety"},{"id":"2609.09604","title":"Watermarks Without Verification: AI Text Watermarking After the EU AI Act","created":"2026-09-09","cited_by_count":0,"area":"safety"},{"id":"2609.19366","title":"The Role of Fine-grained Harm Signals in LLM Safety","created":"2026-09-16","cited_by_count":0,"area":"safety"},{"id":"2609.29948","title":"ENDOPROMPT: Victim-Side Pseudo-References for Utility Degradation","created":"2026-09-24","cited_by_count":0,"area":"safety"},{"id":"2609.32116","title":"Escaping Alignment: A Physical Trap Model of Best-of-N Jailbreaking","created":"2026-09-26","cited_by_count":0,"area":"safety"},{"id":"2609.34514","title":"How to Tame a Multi-Headed Hydra? Adaptive Multi-Category Safety Steering for Large Language Models","created":"2026-09-28","cited_by_count":0,"area":"safety"},{"id":"2609.37858","title":"Storage Is Not Strategy: State-Conditioned Support Control for LLM Unlearning","created":"2026-09-29","cited_by_count":0,"area":"safety"},{"id":"2610.05541","title":"Don't Judge an LLM Only by Its Activations: Discovering Suppressed Safety Features via Counterfactual Activation Potential","created":"2026-10-04","cited_by_count":0,"area":"safety"},{"id":"2610.06503","title":"Harmful Content Generation in Text-to-Image Models: Capabilities and Moderation Limitations","created":"2026-10-05","cited_by_count":0,"area":"safety"},{"id":"2610.07532","title":"Safeguarding LLMs via Model-Agnostic Latent Safety Signals from Dark Knowledge","created":"2026-10-05","cited_by_count":0,"area":"safety"},{"id":"2610.09033","title":"Quad-State Safety Evaluation of Open-Weight Large Language Models on Non-Canonical Inputs","created":"2026-10-06","cited_by_count":0,"area":"safety"},{"id":"2610.17880","title":"Cross-Lingual Transfer of Evil Vectors Is Directionally Asymmetric","created":"2026-10-13","cited_by_count":0,"area":"safety"}],"next_cursor":null}