{"total":93,"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":"2406.18403","title":"LLMs instead of Human Judges? A Large Scale Empirical Study across 20 NLP Evaluation Tasks","created":"2024-06-26","cited_by_count":243,"area":null},{"id":"2407.21772","title":"ShieldGemma: Generative AI Content Moderation Based on Gemma","created":"2024-07-31","cited_by_count":243,"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":"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":"2505.23836","title":"Large Language Models Often Know When They Are Being Evaluated","created":"2025-05-28","cited_by_count":128,"area":null},{"id":"2407.18418","title":"Know Your Limits: A Survey of Abstention in Large Language Models","created":"2024-07-25","cited_by_count":121,"area":null},{"id":"2403.02839","title":"An Empirical Study of LLM-as-a-Judge for LLM Evaluation: Fine-tuned Judge Model is not a General Substitute for GPT-4","created":"2024-03-05","cited_by_count":94,"area":null},{"id":"2311.08370","title":"SimpleSafetyTests: a Test Suite for Identifying Critical Safety Risks in Large Language Models","created":"2023-11-14","cited_by_count":76,"area":null},{"id":"2407.17436","title":"AIR-Bench 2024: A Safety Benchmark Based on Risk Categories from Regulations and Policies","created":"2024-07-11","cited_by_count":74,"area":null},{"id":"2504.04377","title":"PolyGuard: A Multilingual Safety Moderation Tool for 17 Languages","created":"2025-04-06","cited_by_count":59,"area":null},{"id":"2507.16795","title":"Steering Out-of-Distribution Generalization with Concept Ablation Fine-Tuning","created":"2025-07-22","cited_by_count":39,"area":null},{"id":"2502.13175","title":"Towards Robust and Secure Embodied AI: A Survey on Vulnerabilities and Attacks","created":"2025-02-18","cited_by_count":34,"area":null},{"id":"2406.00936","title":"A Survey of Useful LLM Evaluation","created":"2024-06-03","cited_by_count":29,"area":null},{"id":"2504.18838","title":"Toward Generalizable Evaluation in the LLM Era: A Survey Beyond Benchmarks","created":"2025-04-26","cited_by_count":28,"area":null},{"id":"2407.10627","title":"Arena Learning: Build Data Flywheel for LLMs Post-training via Simulated Chatbot Arena","created":"2024-07-15","cited_by_count":26,"area":null},{"id":"2506.11094","title":"The Scales of Justitia: A Comprehensive Survey on Safety Evaluation of LLMs","created":"2025-06-06","cited_by_count":26,"area":null},{"id":"2501.00192","title":"MLLM-as-a-Judge for Image Safety without Human Labeling","created":"2024-12-31","cited_by_count":23,"area":null},{"id":"2503.06550","title":"BingoGuard: LLM Content Moderation Tools with Risk Levels","created":"2025-03-09","cited_by_count":21,"area":null},{"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":"2406.10311","title":"CHiSafetyBench: A Chinese Hierarchical Safety Benchmark for Large Language Models","created":"2024-06-14","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":"2407.08422","title":"On the (In)Security of LLM App Stores","created":"2024-07-11","cited_by_count":14,"area":null},{"id":"2504.12397","title":"Activated LoRA: Fine-tuned LLMs for Intrinsics","created":"2025-04-16","cited_by_count":12,"area":null},{"id":"2410.16665","title":"SafetyAnalyst: Interpretable, Transparent, and Steerable Safety Moderation for AI Behavior","created":"2024-10-22","cited_by_count":10,"area":null},{"id":"2502.11411","title":"Detecting and Filtering Unsafe Training Data via Data Attribution with Denoised Representation","created":"2025-02-17","cited_by_count":9,"area":null},{"id":"2609.37647","title":"Evaluating and Benchmarking the System One Model Jev","created":"2026-09-29","cited_by_count":9,"area":null},{"id":"2404.09785","title":"Benchmarking Llama2, Mistral, Gemma and GPT for Factuality, Toxicity, Bias and Propensity for Hallucinations","created":"2024-04-15","cited_by_count":8,"area":null},{"id":"2504.10277","title":"RealHarm: A Collection of Real-World Language Model Application Failures","created":"2025-04-14","cited_by_count":8,"area":null},{"id":"2412.15035","title":"LLMs Lost in Translation: M-ALERT uncovers Cross-Linguistic Safety Inconsistencies","created":"2024-12-19","cited_by_count":7,"area":null},{"id":"2412.15265","title":"Chinese SafetyQA: A Safety Short-form Factuality Benchmark for Large Language Models","created":"2024-12-17","cited_by_count":7,"area":null},{"id":"2602.23636","title":"FlexGuard: Continuous Risk Scoring for Strictness-Adaptive LLM Content Moderation","created":"2026-02-27","cited_by_count":7,"area":null},{"id":"2410.22307","title":"SVIP: Towards Verifiable Inference of Open-source Large Language Models","created":"2024-10-29","cited_by_count":6,"area":null},{"id":"2506.02945","title":"Quantitative LLM Judges","created":"2025-06-03","cited_by_count":6,"area":null},{"id":"2409.06072","title":"DetoxBench: Benchmarking Large Language Models for Multitask Fraud & Abuse Detection","created":"2024-09-09","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":"2412.21052","title":"Towards Effective Discrimination Testing for Generative AI","created":"2024-12-30","cited_by_count":5,"area":null},{"id":"2507.09424","title":"DATE-LM: Benchmarking Data Attribution Evaluation for Large Language Models","created":"2025-07-12","cited_by_count":5,"area":null},{"id":"2506.07330","title":"JavelinGuard: Low-Cost Transformer Architectures for LLM Security","created":"2025-06-09","cited_by_count":4,"area":null},{"id":"2512.23573","title":"ProGuard: Towards Proactive Multimodal Safeguard","created":"2025-12-29","cited_by_count":4,"area":null},{"id":"2603.05786","title":"Proof-of-Guardrail in AI Agents and What (Not) to Trust from It","created":"2026-03-06","cited_by_count":4,"area":null},{"id":"2605.28591","title":"Models That Know How Evaluations Are Designed Score Safer","created":"2026-05-27","cited_by_count":4,"area":null},{"id":"2409.15766","title":"CHBench: A Chinese Dataset for Evaluating Health in Large Language Models","created":"2024-09-24","cited_by_count":3,"area":null},{"id":"2410.08442","title":"JurEE not Judges: safeguarding llm interactions with small, specialised Encoder Ensembles","created":"2024-10-11","cited_by_count":3,"area":null},{"id":"2509.18293","title":"Evaluating Large Language Models for Detecting Antisemitism","created":"2025-09-22","cited_by_count":3,"area":null},{"id":"2601.13922","title":"Automatic Prompt Optimization for Dataset-Level Feature Discovery","created":"2026-01-20","cited_by_count":3,"area":null},{"id":"2604.17730","title":"MHSafeEval: Role-Aware Interaction-Level Evaluation of Mental Health Safety in Large Language Models","created":"2026-04-20","cited_by_count":3,"area":null},{"id":"2608.11772","title":"Diagnosis Before Recovery: Turning Agent Failures into Selective Self-Correction","created":"2026-08-12","cited_by_count":3,"area":null},{"id":"2501.06137","title":"Supervision policies can shape long-term risk management in general-purpose AI models","created":"2025-01-10","cited_by_count":2,"area":null},{"id":"2506.02326","title":"Something Just Like TRuST : Toxicity Recognition of Span and Target","created":"2025-06-02","cited_by_count":2,"area":null},{"id":"2604.11943","title":"ProbeLogits: Kernel-Level LLM Inference Primitives for AI-Native Operating Systems","created":"2026-04-13","cited_by_count":2,"area":null},{"id":"2604.23099","title":"ProEval: Proactive Failure Discovery and Efficient Performance Estimation for Generative AI Evaluation","created":"2026-04-25","cited_by_count":2,"area":null},{"id":"2605.29427","title":"FinGuard: Detecting Financial Regulatory Non-Compliance in LLM Interactions","created":"2026-05-28","cited_by_count":2,"area":null},{"id":"2608.20350","title":"How to Train a Real-World Silicon Concierge? Internalizing Complex Business Workflow to Only OneModel","created":"2026-06-16","cited_by_count":2,"area":null},{"id":"2608.27147","title":"Thomson: Continual Learning of Frontier Models for SovereignAI","created":"2026-08-27","cited_by_count":2,"area":null},{"id":"2405.19299","title":"Expert-Guided Extinction of Toxic Tokens for Debiased Generation","created":"2024-05-29","cited_by_count":1,"area":null},{"id":"2502.14969","title":"Lost in Space: Finding the Right Tokens for Structured Output","created":"2025-02-20","cited_by_count":1,"area":null},{"id":"2507.20264","title":"EMBRACE: Shaping Inclusive Opinion Representation by Aligning Implicit Conversations with Social Norms","created":"2025-07-27","cited_by_count":1,"area":null},{"id":"2509.06326","title":"AttestLLM: Efficient Attestation Framework for Billion-scale On-device LLMs","created":"2025-09-08","cited_by_count":1,"area":null},{"id":"2512.24991","title":"Efficiently Estimating Data Efficiency for Language Model Fine-tuning","created":"2025-12-31","cited_by_count":1,"area":null},{"id":"2604.07655","title":"Guardian-as-an-Advisor: Advancing Next-Generation Guardian Models for Trustworthy LLMs","created":"2026-04-08","cited_by_count":1,"area":null},{"id":"2604.25580","title":"Bye Bye Perspective API: Lessons for Building and Governing Measurement Infrastructure","created":"2026-04-28","cited_by_count":1,"area":null},{"id":"2605.20203","title":"GrandGuard: Taxonomy, Benchmark, and Safeguards for Elderly-Chatbot Interaction Safety","created":"2026-04-07","cited_by_count":1,"area":null},{"id":"2606.10487","title":"Stop Early, Spend Less: Hidden-State Probes as a Practical Recipe for Streaming Moderation of LLM Outputs","created":"2026-06-09","cited_by_count":1,"area":null},{"id":"2607.02079","title":"HaloGuard 1.0: An Open Weights Constitutional Classifier for Multilingual AI Safety","created":"2026-07-02","cited_by_count":1,"area":null},{"id":"2607.13039","title":"Safeguard-Conditioned Uplift: Measuring Utility-Risk Frontiers for Dual-Use Biology Assistants","created":"2026-06-12","cited_by_count":1,"area":null},{"id":"2607.25857","title":"Shieldstral","created":"2026-07-28","cited_by_count":1,"area":null},{"id":"2609.30706","title":"LAVOIR: Teaching a Single-Pass Decision Encoder When and What to Ask with Amortized Value of Information","created":"2026-09-25","cited_by_count":1,"area":null},{"id":"2502.12411","title":"Gradient Co-occurrence Analysis for Detecting Unsafe Prompts in Large Language Models","created":"2025-02-18","cited_by_count":0,"area":null},{"id":"2506.06391","title":"From Rogue to Safe AI: The Role of Explicit Refusals in Aligning LLMs with International Humanitarian Law","created":"2025-06-05","cited_by_count":0,"area":null},{"id":"2509.09043","title":"Stated Preference for Interaction and Continued Engagement (SPICE): Evaluating an LLM's Willingness to Re-engage in Conversation","created":"2025-09-10","cited_by_count":0,"area":null},{"id":"2509.20097","title":"Integrated Framework for LLM Evaluation with Answer Generation","created":"2025-09-24","cited_by_count":0,"area":null},{"id":"2510.05864","title":"On the Sensitivity of Instruction-tuned LLMs to Harmful Sentences in Long Inputs","created":"2025-10-07","cited_by_count":0,"area":null},{"id":"2601.13288","title":"A BERTology View of LLM Orchestrations: Token- and Layer-Selective Probes for Efficient Single-Pass Classification","created":"2026-01-19","cited_by_count":0,"area":null},{"id":"2603.07286","title":"Taiwan Safety Benchmark and Breeze Guard: Toward Trustworthy AI for Taiwanese Mandarin","created":"2026-03-07","cited_by_count":0,"area":null},{"id":"2604.01745","title":"Detecting Toxic Language: Ontology and BERT-based Approaches for Bulgarian Text","created":"2026-04-02","cited_by_count":0,"area":null},{"id":"2605.00269","title":"Useful Features, Backward Scores: OOD in Language-Model Trajectories","created":"2026-04-30","cited_by_count":0,"area":null},{"id":"2605.10639","title":"Navigating the Sea of LLM Evaluation: Investigating Bias in Toxicity Benchmarks","created":"2026-05-11","cited_by_count":0,"area":null},{"id":"2605.20241","title":"Geometry-Lite: Interpretable Safety Probing via Layer-Wise Margin Geometry","created":"2026-05-18","cited_by_count":0,"area":null},{"id":"2605.25893","title":"$D^2$-Monitor: Dynamic Safety Monitoring for Diffusion LLMs via Hesitation Signals","created":"2026-05-25","cited_by_count":0,"area":null},{"id":"2606.20626","title":"Efficient Safety Benchmarking via Item Response Theory","created":"2026-05-26","cited_by_count":0,"area":null},{"id":"2607.25063","title":"Similar Models Learn Differently: Final-Window Pretraining Shapes Post-Training Beyond SFT","created":"2026-07-27","cited_by_count":0,"area":null},{"id":"2607.26200","title":"Choosing Where and How to Moderate: End-to-End Trade-offs in Filter Placement and Response Rewriting","created":"2026-07-28","cited_by_count":0,"area":null},{"id":"2608.00180","title":"C-Instrument: Automating RL Data Generation and Hillclimbing with a Constitution-Grid Instrument","created":"2026-07-31","cited_by_count":0,"area":null},{"id":"2608.01291","title":"ArabicDialectSafety: A Dialect-Aware Benchmark for Arabic Content Safety Classification","created":"2026-08-02","cited_by_count":0,"area":null},{"id":"2608.11200","title":"ConVAWG: A Retrieval-Grounded Framework for Controlled Synthetic Dialogue Generation in Violence Against Women and Girls","created":"2026-08-11","cited_by_count":0,"area":null},{"id":"2608.13622","title":"ARC: Fair Relative Advantage Comparison in Open-Ended Real-World Interaction","created":"2026-08-13","cited_by_count":0,"area":null},{"id":"2608.21570","title":"A Reproducible, License-Aware Distillation Recipe for CPUDeployable Safety Classification","created":"2026-08-21","cited_by_count":0,"area":null},{"id":"2609.01210","title":"Who Judges the Judges? A Chinese Safety QA Benchmark for Evaluating LLM Responses and Safety Judges","created":"2026-09-01","cited_by_count":0,"area":null},{"id":"2609.01611","title":"EvalDetectBench: A Benchmark for Measuring Evaluation Awareness in Frontier Language Models","created":"2026-06-08","cited_by_count":0,"area":null},{"id":"2609.03814","title":"Evaluating Criterion-Conditioned Behaviour of Large Language Models in Content Moderation","created":"2026-09-03","cited_by_count":0,"area":null},{"id":"2609.32267","title":"When Can Old Evaluations Certify a New Model? Label-Efficient Release Decisions under Evaluator Drift","created":"2026-09-26","cited_by_count":0,"area":null},{"id":"2609.39111","title":"Bongard: Training Machine Intuition","created":"2026-09-30","cited_by_count":0,"area":null}],"next_cursor":null}