{"total":83,"rows":[{"id":"2507.03724","title":"MemOS: A Memory OS for AI System","created":"2025-07-04","cited_by_count":138,"area":"rag"},{"id":"2403.09572","title":"Eyes Closed, Safety On: Protecting Multimodal LLMs via Image-to-Text Transformation","created":"2024-03-14","cited_by_count":124,"area":"lr"},{"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":"2410.20290","title":"Fast Best-of-N Decoding via Speculative Rejection","created":"2024-10-26","cited_by_count":112,"area":null},{"id":"2406.12030","title":"SPA-VL: A Comprehensive Safety Preference Alignment Dataset for Vision Language Model","created":"2024-06-17","cited_by_count":85,"area":"mllm"},{"id":"2407.07403","title":"A Survey of Attacks on Large Vision-Language Models: Resources, Advances, and Future Trends","created":"2024-07-10","cited_by_count":84,"area":"mllm"},{"id":"2402.00357","title":"Safety of Multimodal Large Language Models on Images and Texts","created":"2024-02-01","cited_by_count":75,"area":"mllm"},{"id":"2502.14881","title":"A Survey of Safety on Large Vision-Language Models: Attacks, Defenses and Evaluations","created":"2025-02-14","cited_by_count":67,"area":"mllm"},{"id":"2504.17704","title":"Safety in Large Reasoning Models: A Survey","created":"2025-04-24","cited_by_count":64,"area":"lr"},{"id":"2411.19939","title":"VLSBench: Unveiling Visual Leakage in Multimodal Safety","created":"2024-11-29","cited_by_count":63,"area":"mllm"},{"id":"2411.16721","title":"Steering Away from Harm: An Adaptive Approach to Defending Vision Language Model Against Jailbreaks","created":"2024-11-23","cited_by_count":48,"area":"mllm"},{"id":"2404.05600","title":"SpeechAlign: Aligning Speech Generation to Human Preferences","created":"2024-04-08","cited_by_count":47,"area":"alignment"},{"id":"2501.01765","title":"SaLoRA: Safety-Alignment Preserved Low-Rank Adaptation","created":"2025-01-03","cited_by_count":42,"area":"safety"},{"id":"2505.11049","title":"GuardReasoner-VL: Safeguarding VLMs via Reinforced Reasoning","created":"2025-05-16","cited_by_count":42,"area":"lr"},{"id":"2504.09466","title":"AdaSteer: Your Aligned LLM is Inherently an Adaptive Jailbreak Defender","created":"2025-04-13","cited_by_count":38,"area":"alignment"},{"id":"2411.09259","title":"Jailbreak Attacks and Defenses against Multimodal Generative Models: A Survey","created":"2024-11-14","cited_by_count":32,"area":"safety"},{"id":"2401.16332","title":"Tradeoffs Between Alignment and Helpfulness in Language Models with Steering Methods","created":"2024-01-29","cited_by_count":28,"area":"alignment"},{"id":"2406.15279","title":"Safe Inputs but Unsafe Output: Benchmarking Cross-modality Safety Alignment of Large Vision-Language Model","created":"2024-06-21","cited_by_count":26,"area":"mllm"},{"id":"2407.21659","title":"Cross-modality Information Check for Detecting Jailbreaking in Multimodal Large Language Models","created":"2024-07-31","cited_by_count":25,"area":"mllm"},{"id":"2410.09047","title":"Unraveling and Mitigating Safety Alignment Degradation of Vision-Language Models","created":"2024-10-11","cited_by_count":23,"area":"mllm"},{"id":"2410.10343","title":"Locking Down the Finetuned LLMs Safety","created":"2024-10-14","cited_by_count":22,"area":null},{"id":"2411.08410","title":"The VLLM Safety Paradox: Dual Ease in Jailbreak Attack and Defense","created":"2024-11-13","cited_by_count":21,"area":"safety"},{"id":"2502.17601","title":"Representation Engineering for Large-Language Models: Survey and Research Challenges","created":"2025-02-24","cited_by_count":21,"area":"mechinterp"},{"id":"2404.05264","title":"Unbridled Icarus: A Survey of the Potential Perils of Image Inputs in Multimodal Large Language Model Security","created":"2024-04-08","cited_by_count":19,"area":"mllm"},{"id":"2402.19465","title":"Towards Tracing Trustworthiness Dynamics: Revisiting Pre-training Period of Large Language Models","created":"2024-02-29","cited_by_count":18,"area":null},{"id":"2410.01174","title":"Towards Inference-time Category-wise Safety Steering for Large Language Models","created":"2024-10-02","cited_by_count":18,"area":"alignment"},{"id":"2502.10486","title":"VLM-Guard: Safeguarding Vision-Language Models via Fulfilling Safety Alignment Gap","created":"2025-02-14","cited_by_count":17,"area":"mllm"},{"id":"2502.13095","title":"Understanding and Rectifying Safety Perception Distortion in VLMs","created":"2025-02-18","cited_by_count":17,"area":"mllm"},{"id":"2507.13255","title":"Automating Steering for Safe Multimodal Large Language Models","created":"2025-07-17","cited_by_count":17,"area":"lr"},{"id":"2503.11832","title":"Safety Mirage: How Spurious Correlations Undermine VLM Safety Fine-Tuning and Can Be Mitigated by Machine Unlearning","created":"2025-03-14","cited_by_count":16,"area":"mllm"},{"id":"2502.01208","title":"On Almost Surely Safe Alignment of Large Language Models at Inference-Time","created":"2025-02-03","cited_by_count":14,"area":"alignment"},{"id":"2411.08733","title":"Dynamic Rewarding with Prompt Optimization Enables Tuning-free Self-Alignment of Language Models","created":"2024-11-13","cited_by_count":13,"area":"alignment"},{"id":"2509.13281","title":"RepIt: Steering Language Models with Concept-Specific Refusal Vectors","created":"2025-09-16","cited_by_count":12,"area":null},{"id":"2412.00142","title":"Enhancing Few-Shot Vision-Language Classification with Large Multimodal Model Features","created":"2024-11-28","cited_by_count":11,"area":"mllm"},{"id":"2506.04250","title":"SafeSteer: Interpretable Safety Steering with Refusal-Evasion in LLMs","created":"2025-06-01","cited_by_count":10,"area":"mechinterp"},{"id":"2406.01514","title":"Decoupled Alignment for Robust Plug-and-Play Adaptation","created":"2024-06-03","cited_by_count":9,"area":"alignment"},{"id":"2406.18118","title":"SafeAligner: Safety Alignment against Jailbreak Attacks via Response Disparity Guidance","created":"2024-06-26","cited_by_count":9,"area":"alignment"},{"id":"2505.12050","title":"AdaBoN: Adaptive Best-of-N Alignment","created":"2025-05-17","cited_by_count":9,"area":"alignment"},{"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":"2506.09353","title":"DAVSP: Safety Alignment for Large Vision-Language Models via Deep Aligned Visual Safety Prompt","created":"2025-06-11","cited_by_count":8,"area":"mllm"},{"id":"2604.14090","title":"From Weights to Activations: Is Steering the Next Frontier of Adaptation?","created":"2026-04-15","cited_by_count":8,"area":null},{"id":"2505.20081","title":"Inference-time Alignment in Continuous Space","created":"2025-05-26","cited_by_count":7,"area":"alignment"},{"id":"2602.10161","title":"Omni-Safety under Cross-Modality Conflict: Vulnerabilities, Dynamics Mechanisms and Efficient Alignment","created":"2026-02-10","cited_by_count":6,"area":null},{"id":"2406.08124","title":"Legend: Leveraging Representation Engineering to Annotate Safety Margin for Preference Datasets","created":"2024-06-12","cited_by_count":5,"area":"alignment"},{"id":"2410.14184","title":"MetaAlign: Align Large Language Models with Diverse Preferences during Inference Time","created":"2024-10-18","cited_by_count":5,"area":"alignment"},{"id":"2509.00373","title":"Activation Steering Meets Preference Optimization: Defense Against Jailbreaks in Vision Language Models","created":"2025-08-30","cited_by_count":5,"area":"mllm"},{"id":"2509.12934","title":"The Anatomy of Alignment: Decomposing Preference Optimization by Steering Sparse Features","created":"2025-09-16","cited_by_count":5,"area":"alignment"},{"id":"2604.08169","title":"Activation Steering for Aligned Open-ended Generation without Sacrificing Coherence","created":"2026-04-09","cited_by_count":4,"area":"safety"},{"id":"2605.15030","title":"WARD: Adversarially Robust Defense of Web Agents Against Prompt Injections","created":"2026-05-14","cited_by_count":4,"area":"agents"},{"id":"2408.08464","title":"$\\textit{MMJ-Bench}$: A Comprehensive Study on Jailbreak Attacks and Defenses for Multimodal Large Language Models","created":"2024-08-16","cited_by_count":3,"area":"safety"},{"id":"2505.24208","title":"Bootstrapping LLM Robustness for VLM Safety via Reducing the Pretraining Modality Gap","created":"2025-05-30","cited_by_count":3,"area":"mllm"},{"id":"2601.03537","title":"STAR-S: Improving Safety Alignment through Self-Taught Reasoning on Safety Rules","created":"2026-01-07","cited_by_count":3,"area":"lr"},{"id":"2601.03594","title":"Jailbreaking LLMs & VLMs: Mechanisms, Evaluation, and Unified Defense","created":"2026-01-07","cited_by_count":3,"area":"safety"},{"id":"2602.07013","title":"Steering to Say No: Configurable Refusal via Activation Steering in Vision Language Models","created":"2026-01-31","cited_by_count":3,"area":"mllm"},{"id":"2410.19206","title":"Inference time LLM alignment in single and multidomain preference spectrum","created":"2024-10-24","cited_by_count":2,"area":"alignment"},{"id":"2502.20795","title":"Test-Time Alignment for Large Language Models via Textual Model Predictive Control","created":"2025-02-28","cited_by_count":2,"area":"alignment"},{"id":"2508.09016","title":"A Survey on Training-free Alignment of Large Language Models","created":"2025-08-12","cited_by_count":2,"area":"alignment"},{"id":"2509.04403","title":"Self-adaptive Dataset Construction for Real-World Multimodal Safety Scenarios","created":"2025-09-04","cited_by_count":2,"area":"mllm"},{"id":"2511.16324","title":"SDA: Steering-Driven Distribution Alignment for Open LLMs without Fine-Tuning","created":"2025-11-20","cited_by_count":2,"area":"alignment"},{"id":"2512.19456","title":"Activations as Features: Probing LLMs for Generalizable Essay Scoring Representations","created":"2025-12-22","cited_by_count":2,"area":null},{"id":"2602.03402","title":"Risk Awareness Injection: Calibrating Vision-Language Models for Safety without Compromising Utility","created":"2026-02-03","cited_by_count":2,"area":"lr"},{"id":"2604.00012","title":"Finding and Reactivating Post-Trained LLMs' Hidden Safety Mechanisms","created":"2026-03-10","cited_by_count":2,"area":"lr"},{"id":"2604.26167","title":"Test-Time Safety Alignment","created":"2026-04-28","cited_by_count":2,"area":"alignment"},{"id":"2606.11201","title":"To Intervene or Not: Guiding Inference-time Alignment with Probabilistic Model Blending","created":"2026-04-22","cited_by_count":2,"area":"alignment"},{"id":"2607.09697","title":"Safe responses matter: Output-aware safety guardrail mitigate over-refusal in MLLMs","created":"2026-06-20","cited_by_count":2,"area":"mllm"},{"id":"2506.13774","title":"Personalized Constitutionally-Aligned Agentic Superego: Secure AI Behavior Aligned to Diverse Human Values","created":"2025-06-08","cited_by_count":1,"area":"agents"},{"id":"2511.14195","title":"N-GLARE: An Non-Generative Latent Representation-Efficient LLM Safety Evaluator","created":"2025-11-18","cited_by_count":1,"area":"safety"},{"id":"2602.07107","title":"ShallowJail: Steering Jailbreaks against Large Language Models","created":"2026-02-06","cited_by_count":1,"area":"safety"},{"id":"2603.14825","title":"Two Birds, One Projection: Harmonizing Safety and Utility in LVLMs via Inference-time Feature Projection","created":"2026-03-16","cited_by_count":1,"area":"mllm"},{"id":"2608.30319","title":"Beyond Token-Level Guidance: Inference-Time Alignment of Specialized LLMs via Cross-Family Representation Steering","created":"2026-08-31","cited_by_count":1,"area":"mechinterp"},{"id":"2609.00092","title":"Safin-1: Safety from Within through Memory-Native State Evolution","created":"2026-08-31","cited_by_count":1,"area":null},{"id":"2509.21400","title":"SafeSteer: Adaptive Subspace Steering for Efficient Jailbreak Defense in Vision-Language Models","created":"2025-09-24","cited_by_count":0,"area":"safety"},{"id":"2510.09330","title":"Safety Game: Inference-Time Alignment of Black-Box LLMs via Constrained Optimization","created":"2025-10-10","cited_by_count":0,"area":"alignment"},{"id":"2602.17778","title":"Asking Forever: Universal Activations Behind Turn Amplification in Conversational LLMs","created":"2026-02-19","cited_by_count":0,"area":null},{"id":"2604.15789","title":"A Systematic Study of Training-Free Methods for Trustworthy Large Language Models","created":"2026-04-17","cited_by_count":0,"area":null},{"id":"2604.17210","title":"Guardrails in Logit Space: Safety Token Regularization for LLM Alignment","created":"2026-04-19","cited_by_count":0,"area":"alignment"},{"id":"2608.07535","title":"Evolving Safety Landscape of Multi-modal Large Language Models: A Survey of Emerging Threats and Safeguards","created":"2026-07-27","cited_by_count":0,"area":"mllm"},{"id":"2608.14797","title":"Beyond Tokens: A Survey on Decoding Methods for Large Language and Vision-Language Models","created":"2026-08-14","cited_by_count":0,"area":"mllm"},{"id":"2609.03544","title":"SafeRI: Recognition and Intervention for Token-Level Safety Intervention in Large Vision Language Models","created":"2026-09-03","cited_by_count":0,"area":"mllm"},{"id":"2609.04022","title":"Representational alignment yields generalizable safety in language models","created":"2026-09-03","cited_by_count":0,"area":"alignment"},{"id":"2609.14258","title":"SPARK: Representation-Level KV Memory Alignment for Safer Vision-Language Models","created":"2026-09-13","cited_by_count":0,"area":"mllm"},{"id":"2609.19325","title":"AUDITPLAN: Commit, Then Answer for Auditable Safety Alignment","created":"2026-09-16","cited_by_count":0,"area":"alignment"},{"id":"2610.19817","title":"Model vs. User: Whose Values Prevail? Inference-Time Value Alignment in Multi-Turn Conversations","created":"2026-10-14","cited_by_count":0,"area":"alignment"}],"next_cursor":null}