{"total":185,"rows":[{"id":"2312.06674","title":"Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations","created":"2023-12-07","cited_by_count":1207,"area":null},{"id":"2310.03693","title":"Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!","created":"2023-10-05","cited_by_count":1154,"area":"alignment"},{"id":"2307.04657","title":"BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Dataset","created":"2023-07-10","cited_by_count":860,"area":"alignment"},{"id":"2406.18495","title":"WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs","created":"2024-06-26","cited_by_count":496,"area":"safety"},{"id":"2208.03274","title":"A Holistic Approach to Undesired Content Detection in the Real World","created":"2022-08-05","cited_by_count":356,"area":null},{"id":"2205.05131","title":"UL2: Unifying Language Learning Paradigms","created":"2022-05-10","cited_by_count":295,"area":null},{"id":"2305.18565","title":"PaLI-X: On Scaling up a Multilingual Vision and Language Model","created":"2023-05-29","cited_by_count":255,"area":null},{"id":"2402.05044","title":"SALAD-Bench: A Hierarchical and Comprehensive Safety Benchmark for Large Language Models","created":"2024-02-07","cited_by_count":228,"area":null},{"id":"2412.03555","title":"PaliGemma 2: A Family of Versatile VLMs for Transfer","created":"2024-12-04","cited_by_count":207,"area":"mllm"},{"id":"2305.13169","title":"A Pretrainer's Guide to Training Data: Measuring the Effects of Data Age, Domain Coverage, Quality, & Toxicity","created":"2023-05-22","cited_by_count":182,"area":null},{"id":"2404.05993","title":"AEGIS: Online Adaptive AI Content Safety Moderation with Ensemble of LLM Experts","created":"2024-04-09","cited_by_count":136,"area":null},{"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":"2310.09199","title":"PaLI-3 Vision Language Models: Smaller, Faster, Stronger","created":"2023-10-13","cited_by_count":126,"area":"mllm"},{"id":"2406.09187","title":"GuardAgent: Safeguard LLM Agents by a Guard Agent via Knowledge-Enabled Reasoning","created":"2024-06-13","cited_by_count":124,"area":"lr"},{"id":"2501.18492","title":"GuardReasoner: Towards Reasoning-based LLM Safeguards","created":"2025-01-30","cited_by_count":100,"area":"lr"},{"id":"2405.14577","title":"Representation Noising: A Defence Mechanism Against Harmful Finetuning","created":"2024-05-23","cited_by_count":96,"area":"safety"},{"id":"2303.11504","title":"Language Model Behavior: A Comprehensive Survey","created":"2023-03-20","cited_by_count":92,"area":null},{"id":"2402.06627","title":"Feedback Loops With Language Models Drive In-Context Reward Hacking","created":"2024-02-09","cited_by_count":77,"area":"alignment"},{"id":"2403.13031","title":"RigorLLM: Resilient Guardrails for Large Language Models against Undesired Content","created":"2024-03-19","cited_by_count":68,"area":"safety"},{"id":"2404.01349","title":"Fairness in Large Language Models: A Taxonomic Survey","created":"2024-03-31","cited_by_count":58,"area":null},{"id":"2402.16444","title":"ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors","created":"2024-02-26","cited_by_count":51,"area":null},{"id":"2404.12670","title":"Towards Human-centered Proactive Conversational Agents","created":"2024-04-19","cited_by_count":49,"area":null},{"id":"2304.12397","title":"On the Challenges of Using Black-Box APIs for Toxicity Evaluation in Research","created":"2023-04-24","cited_by_count":44,"area":null},{"id":"2407.05557","title":"$R^2$-Guard: Robust Reasoning Enabled LLM Guardrail via Knowledge-Enhanced Logical Reasoning","created":"2024-07-08","cited_by_count":43,"area":"lr"},{"id":"2310.05280","title":"Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systems","created":"2023-10-08","cited_by_count":37,"area":null},{"id":"2306.07377","title":"Lost in Translation: Large Language Models in Non-English Content Analysis","created":"2023-06-12","cited_by_count":35,"area":null},{"id":"2401.16656","title":"Gradient-Based Language Model Red Teaming","created":"2024-01-30","cited_by_count":34,"area":"safety"},{"id":"2407.03963","title":"LLM-jp: A Cross-organizational Project for the Research and Development of Fully Open Japanese LLMs","created":"2024-07-04","cited_by_count":33,"area":null},{"id":"2406.16235","title":"Preference Tuning For Toxicity Mitigation Generalizes Across Languages","created":"2024-06-23","cited_by_count":31,"area":"alignment"},{"id":"2404.01322","title":"A Review of Multi-Modal Large Language and Vision Models","created":"2024-03-28","cited_by_count":30,"area":null},{"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":"2311.04124","title":"Unveiling Safety Vulnerabilities of Large Language Models","created":"2023-11-07","cited_by_count":27,"area":null},{"id":"2312.11523","title":"ToViLaG: Your Visual-Language Generative Model is Also An Evildoer","created":"2023-12-13","cited_by_count":27,"area":null},{"id":"2311.01270","title":"People Make Better Edits: Measuring the Efficacy of LLM-Generated Counterfactually Augmented Data for Harmful Language Detection","created":"2023-11-02","cited_by_count":25,"area":null},{"id":"2405.05378","title":"\"They are uncultured\": Unveiling Covert Harms and Social Threats in LLM Generated Conversations","created":"2024-05-08","cited_by_count":25,"area":null},{"id":"2405.11070","title":"Jill Watson: A Virtual Teaching Assistant powered by ChatGPT","created":"2024-05-17","cited_by_count":24,"area":null},{"id":"2410.10414","title":"On Calibration of LLM-based Guard Models for Reliable Content Moderation","created":"2024-10-14","cited_by_count":23,"area":"safety"},{"id":"2410.23687","title":"Adversarial Attacks of Vision Tasks in the Past 10 Years: A Survey","created":"2024-10-31","cited_by_count":23,"area":"mllm"},{"id":"2407.07342","title":"Multilingual Blending: LLM Safety Alignment Evaluation with Language Mixture","created":"2024-07-10","cited_by_count":22,"area":"alignment"},{"id":"2410.03730","title":"Teuken-7B-Base & Teuken-7B-Instruct: Towards European LLMs","created":"2024-09-30","cited_by_count":22,"area":null},{"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":"2504.14904","title":"VLM as Policy: Common-Law Content Moderation Framework for Short Video Platform","created":"2025-04-21","cited_by_count":21,"area":null},{"id":"2506.10597","title":"SoK: Evaluating Jailbreak Guardrails for Large Language Models","created":"2025-06-12","cited_by_count":21,"area":"alignment"},{"id":"2405.09373","title":"PolygloToxicityPrompts: Multilingual Evaluation of Neural Toxic Degeneration in Large Language Models","created":"2024-05-15","cited_by_count":20,"area":null},{"id":"2403.03893","title":"From One to Many: Expanding the Scope of Toxicity Mitigation in Language Models","created":"2024-03-06","cited_by_count":18,"area":null},{"id":"2312.01700","title":"Data Management For Training Large Language Models: A Survey","created":"2023-12-04","cited_by_count":17,"area":null},{"id":"2408.03573","title":"AcTracer: Active Testing of Large Language Model via Multi-Stage Sampling","created":"2024-08-07","cited_by_count":17,"area":null},{"id":"2406.00020","title":"Harmful Speech Detection by Language Models Exhibits Gender-Queer Dialect Bias","created":"2024-05-23","cited_by_count":16,"area":null},{"id":"2411.11504","title":"Search, Verify and Feedback: Towards Next Generation Post-training Paradigm of Foundation Models via Verifier Engineering","created":"2024-11-18","cited_by_count":16,"area":null},{"id":"2310.15431","title":"What Makes it Ok to Set a Fire? Iterative Self-distillation of Contexts and Rationales for Disambiguating Defeasible Social and Moral Situations","created":"2023-10-24","cited_by_count":15,"area":null},{"id":"2412.07338","title":"Contextualized Counterspeech: Strategies for Adaptation, Personalization, and Evaluation","created":"2024-12-10","cited_by_count":15,"area":null},{"id":"2406.01364","title":"BELLS: A Framework Towards Future Proof Benchmarks for the Evaluation of LLM Safeguards","created":"2024-06-03","cited_by_count":14,"area":null},{"id":"2305.17696","title":"SQuARe: A Large-Scale Dataset of Sensitive Questions and Acceptable Responses Created Through Human-Machine Collaboration","created":"2023-05-28","cited_by_count":13,"area":null},{"id":"2308.04813","title":"CLEVA: Chinese Language Models EVAluation Platform","created":"2023-08-09","cited_by_count":13,"area":null},{"id":"2408.11727","title":"Efficient Detection of Toxic Prompts in Large Language Models","created":"2024-08-21","cited_by_count":13,"area":null},{"id":"2411.15462","title":"HateDay: Insights from a Global Hate Speech Dataset Representative of a Day on Twitter","created":"2024-11-23","cited_by_count":13,"area":null},{"id":"2412.04139","title":"Monet: Mixture of Monosemantic Experts for Transformers","created":"2024-12-05","cited_by_count":13,"area":"mechinterp"},{"id":"2406.16746","title":"The Responsible Foundation Model Development Cheatsheet: A Review of Tools & Resources","created":"2024-06-24","cited_by_count":11,"area":null},{"id":"2407.01376","title":"Badllama 3: removing safety finetuning from Llama 3 in minutes","created":"2024-07-01","cited_by_count":11,"area":"safety"},{"id":"2311.07879","title":"Toxicity Detection is NOT all you Need: Measuring the Gaps to Supporting Volunteer Content Moderators","created":"2023-11-14","cited_by_count":10,"area":null},{"id":"2312.10269","title":"The DSA Transparency Database: Auditing Self-reported Moderation Actions by Social Media","created":"2023-12-16","cited_by_count":10,"area":null},{"id":"2403.19260","title":"NaijaHate: Evaluating Hate Speech Detection on Nigerian Twitter Using Representative Data","created":"2024-03-28","cited_by_count":10,"area":null},{"id":"2411.11407","title":"The Dark Side of Trust: Authority Citation-Driven Jailbreak Attacks on Large Language Models","created":"2024-11-18","cited_by_count":10,"area":"safety"},{"id":"2603.19649","title":"PolicySim: An LLM-Based Agent Social Simulation Sandbox for Proactive Policy Optimization","created":"2026-03-20","cited_by_count":10,"area":"agents"},{"id":"2410.19406","title":"An Auditing Test To Detect Behavioral Shift in Language Models","created":"2024-10-25","cited_by_count":9,"area":null},{"id":"2504.10286","title":"Characterizing LLM-driven Social Network: The Chirper.ai Case","created":"2025-04-14","cited_by_count":9,"area":"agents"},{"id":"2507.00657","title":"Generative Exaggeration in LLM Social Agents: Consistency, Bias, and Toxicity","created":"2025-07-01","cited_by_count":9,"area":"agents"},{"id":"2408.09326","title":"Characterizing and Evaluating the Reliability of LLMs against Jailbreak Attacks","created":"2024-08-18","cited_by_count":8,"area":"safety"},{"id":"2411.14398","title":"Lightweight Safety Guardrails Using Fine-tuned BERT Embeddings","created":"2024-11-21","cited_by_count":8,"area":"safety"},{"id":"2511.12710","title":"Evolve the Method, Not the Prompts: Evolutionary Synthesis of Jailbreak Attacks on LLMs","created":"2025-11-16","cited_by_count":8,"area":"safety"},{"id":"2306.00374","title":"CFL: Causally Fair Language Models Through Token-level Attribute Controlled Generation","created":"2023-06-01","cited_by_count":7,"area":null},{"id":"2506.01790","title":"IF-GUIDE: Influence Function-Guided Detoxification of LLMs","created":"2025-06-02","cited_by_count":7,"area":null},{"id":"2301.01874","title":"Critical Perspectives: A Benchmark Revealing Pitfalls in PerspectiveAPI","created":"2023-01-05","cited_by_count":6,"area":null},{"id":"2307.02763","title":"Your spouse needs professional help: Determining the Contextual Appropriateness of Messages through Modeling Social Relationships","created":"2023-07-06","cited_by_count":6,"area":null},{"id":"2404.14695","title":"MisgenderMender: A Community-Informed Approach to Interventions for Misgendering","created":"2024-04-23","cited_by_count":6,"area":null},{"id":"2505.16765","title":"Hiding in Plain Sight: A Steganographic Approach to Stealthy LLM Jailbreaks","created":"2025-05-22","cited_by_count":6,"area":"safety"},{"id":"2506.00166","title":"Disentangled Safety Adapters Enable Efficient Guardrails and Flexible Inference-Time Alignment","created":"2025-05-30","cited_by_count":6,"area":"alignment"},{"id":"2509.08653","title":"Generative Data Refinement: Just Ask for Better Data","created":"2025-09-10","cited_by_count":6,"area":null},{"id":"2509.18127","title":"Safe-SAIL: Towards a Fine-grained Safety Landscape of Large Language Models via Sparse Autoencoder Interpretation Framework","created":"2025-09-11","cited_by_count":6,"area":"mechinterp"},{"id":"2305.12542","title":"ToxBuster: In-game Chat Toxicity Buster with BERT","created":"2023-05-21","cited_by_count":5,"area":null},{"id":"2306.15363","title":"Your Attack Is Too DUMB: Formalizing Attacker Scenarios for Adversarial Transferability","created":"2023-06-27","cited_by_count":5,"area":null},{"id":"2310.18330","title":"Towards Detecting Contextual Real-Time Toxicity for In-Game Chat","created":"2023-10-20","cited_by_count":5,"area":null},{"id":"2406.12800","title":"Supporting Human Raters with the Detection of Harmful Content using Large Language Models","created":"2024-06-18","cited_by_count":5,"area":"safety"},{"id":"2406.17566","title":"FrenchToxicityPrompts: a Large Benchmark for Evaluating and Mitigating Toxicity in French Texts","created":"2024-06-25","cited_by_count":5,"area":null},{"id":"2407.10995","title":"LionGuard: Building a Contextualized Moderation Classifier to Tackle Localized Unsafe Content","created":"2024-06-24","cited_by_count":5,"area":null},{"id":"2409.07772","title":"Alignment with Preference Optimization Is All You Need for LLM Safety","created":"2024-09-12","cited_by_count":5,"area":"alignment"},{"id":"2409.18203","title":"Policy Maps: Tools for Guiding the Unbounded Space of LLM Behaviors","created":"2024-09-26","cited_by_count":5,"area":null},{"id":"2412.13698","title":"Towards Efficient and Explainable Hate Speech Detection via Model Distillation","created":"2024-12-18","cited_by_count":5,"area":null},{"id":"2412.14050","title":"Cross-Lingual Transfer of Debiasing and Detoxification in Multilingual LLMs: An Extensive Investigation","created":"2024-12-18","cited_by_count":5,"area":null},{"id":"2506.06347","title":"Unified Game Moderation: Soft-Prompting and LLM-Assisted Label Transfer for Resource-Efficient Toxicity Detection","created":"2025-06-01","cited_by_count":5,"area":null},{"id":"2510.10998","title":"ABLEIST: Intersectional Disability Bias in LLM-Generated Hiring Scenarios","created":"2025-10-13","cited_by_count":5,"area":null},{"id":"2601.19768","title":"GAVEL: Towards Rule-Based Safety Through Activation Monitoring","created":"2026-01-27","cited_by_count":5,"area":null},{"id":"2210.00910","title":"Hypothesis Engineering for Zero-Shot Hate Speech Detection","created":"2022-10-03","cited_by_count":4,"area":null},{"id":"2311.04345","title":"A Taxonomy of Rater Disagreements: Surveying Challenges & Opportunities from the Perspective of Annotating Online Toxicity","created":"2023-11-07","cited_by_count":4,"area":null},{"id":"2312.15478","title":"A Group Fairness Lens for Large Language Models","created":"2023-12-24","cited_by_count":4,"area":null},{"id":"2406.10847","title":"TorchOpera: A Compound AI System for LLM Safety","created":"2024-06-16","cited_by_count":4,"area":"safety"},{"id":"2411.04291","title":"Layer-wise Alignment: Examining Safety Alignment Across Image Encoder Layers in Vision Language Models","created":"2024-11-06","cited_by_count":4,"area":"mllm"},{"id":"2501.12640","title":"Toxicity Begets Toxicity: Unraveling Conversational Chains in Political Podcasts","created":"2025-01-22","cited_by_count":4,"area":null},{"id":"2502.06394","title":"SynthDetoxM: Modern LLMs are Few-Shot Parallel Detoxification Data Annotators","created":"2025-02-10","cited_by_count":4,"area":null},{"id":"2511.07482","title":"Alignment-Constrained Dynamic Pruning for LLMs: Identifying and Preserving Alignment-Critical Circuits","created":"2025-11-09","cited_by_count":4,"area":null}],"next_cursor":"100"}