{"total":53,"rows":[{"id":"2504.15585","title":"A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment","created":"2025-04-22","cited_by_count":116,"area":"lr"},{"id":"2503.03480","title":"SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Constrained Learning","created":"2025-03-05","cited_by_count":72,"area":"vla"},{"id":"2406.14477","title":"SafeSora: Towards Safety Alignment of Text2Video Generation via a Human Preference Dataset","created":"2024-06-20","cited_by_count":35,"area":"alignment"},{"id":"2509.20758","title":"SFT Doesn't Always Hurt General Capabilities: Revisiting Domain-Specific Fine-Tuning in LLMs","created":"2025-09-25","cited_by_count":24,"area":null},{"id":"2506.05346","title":"Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets","created":"2025-06-05","cited_by_count":20,"area":"alignment"},{"id":"2506.11930","title":"Feedback Friction: LLMs Struggle to Fully Incorporate External Feedback","created":"2025-06-13","cited_by_count":11,"area":null},{"id":"2505.12716","title":"Shadow-FT: Tuning Instruct Model via Training on Paired Base Model","created":"2025-05-19","cited_by_count":10,"area":null},{"id":"2603.24543","title":"Analysing the Safety Pitfalls of Steering Vectors","created":"2026-03-25","cited_by_count":10,"area":"safety"},{"id":"2605.12798","title":"Emergent and Subliminal Misalignment Through the Lens of Data-Mediated Transfer","created":"2026-05-12","cited_by_count":8,"area":null},{"id":"2505.17519","title":"Chain-of-Lure: A Universal Jailbreak Attack Framework using Unconstrained Synthetic Narratives","created":"2025-05-23","cited_by_count":5,"area":"safety"},{"id":"2605.10310","title":"Positive Alignment: Artificial Intelligence for Human Flourishing","created":"2026-05-11","cited_by_count":5,"area":null},{"id":"2602.08159","title":"The Confidence Manifold: Geometric Structure of Correctness Representations in Language Models","created":"2026-02-08","cited_by_count":4,"area":"mechinterp"},{"id":"2603.23268","title":"SafeSeek: Universal Attribution of Safety Circuits in Language Models","created":"2026-03-24","cited_by_count":4,"area":"safety"},{"id":"2604.13076","title":"Alignment midtraining for animals","created":"2026-03-21","cited_by_count":4,"area":null},{"id":"2604.17215","title":"Continual Safety Alignment via Gradient-Based Sample Selection","created":"2026-04-19","cited_by_count":4,"area":"alignment"},{"id":"2607.26654","title":"Constitutional Midtraining: Content Presence Drives Alignment Gains","created":"2026-07-29","cited_by_count":4,"area":null},{"id":"2601.03858","title":"What Does Loss Optimization Actually Teach, If Anything? Knowledge Dynamics in Continual Pre-training of LLMs","created":"2026-01-07","cited_by_count":3,"area":null},{"id":"2601.04448","title":"Merging Triggers, Breaking Backdoors: Defensive Poisoning for Instruction-Tuned Language Models","created":"2026-01-07","cited_by_count":3,"area":"safety"},{"id":"2605.04431","title":"Towards Robust LLM Post-Training: Automatic Failure Management for Reinforcement Fine-Tuning","created":"2026-05-06","cited_by_count":3,"area":null},{"id":"2607.25614","title":"MemSFT: Mitigating Alignment Tax with an External Parametric Memory","created":"2026-07-28","cited_by_count":3,"area":"alignment"},{"id":"2501.05336","title":"Stream Aligner: Efficient Sentence-Level Alignment via Distribution Induction","created":"2025-01-09","cited_by_count":2,"area":"alignment"},{"id":"2507.14198","title":"Retention analysis of edited knowledge after fine-tuning","created":"2025-07-14","cited_by_count":2,"area":null},{"id":"2509.23595","title":"Timber: Training-free Instruct Model Refining with Base via Effective Rank","created":"2025-09-28","cited_by_count":2,"area":null},{"id":"2510.27206","title":"Fints: Efficient Inference-Time Personalization for LLMs with Fine-Grained Instance-Tailored Steering","created":"2025-10-31","cited_by_count":2,"area":null},{"id":"2511.06148","title":"Large Language Models Develop Novel Social Biases Through Adaptive Exploration","created":"2025-11-08","cited_by_count":2,"area":null},{"id":"2512.11912","title":"Robustness of Probabilistic Models to Low-Quality Data: A Multi-Perspective Analysis","created":"2025-12-11","cited_by_count":2,"area":null},{"id":"2603.03081","title":"TAO-Attack: Toward Advanced Optimization-Based Jailbreak Attacks for Large Language Models","created":"2026-03-03","cited_by_count":2,"area":"alignment"},{"id":"2603.10243","title":"GR-SAP: Generative Replay for Safety Alignment Preservation during Fine-Tuning","created":"2026-03-10","cited_by_count":2,"area":"alignment"},{"id":"2605.18309","title":"Alignment Dynamics in LLM Fine-Tuning","created":"2026-05-18","cited_by_count":2,"area":"alignment"},{"id":"2607.09053","title":"An Emergent Mirage: Is Emergent Misalignment and Realignment Indeed a Robust Phenomenon?","created":"2026-07-10","cited_by_count":2,"area":null},{"id":"2504.11830","title":"Emergence of Computational Structure in a Neural Network Physics Simulator","created":"2025-04-16","cited_by_count":1,"area":null},{"id":"2510.11824","title":"Empirical Study on Robustness and Resilience in Cooperative Multi-Agent Reinforcement Learning","created":"2025-10-13","cited_by_count":1,"area":null},{"id":"2512.03704","title":"Overcoming State Inertia: Minimally Invasive Temporal Alignment for Evolving Contexts","created":"2025-12-03","cited_by_count":1,"area":"alignment"},{"id":"2601.05019","title":"H\\'an D\\=an Xu\\'e B\\`u (Mimicry) or Q\\=ing Ch\\=u Y\\'u L\\'an (Mastery)? A Cognitive Perspective on Reasoning Distillation in Large Language Models","created":"2026-01-08","cited_by_count":1,"area":"lr"},{"id":"2604.05995","title":"The Model Agreed, But Didn't Learn: Diagnosing Surface Compliance in Large Language Models","created":"2026-04-07","cited_by_count":1,"area":null},{"id":"2606.19168","title":"Beyond Safe Data: Pretraining-Stage Alignment with Regular Safety Reflection","created":"2026-06-17","cited_by_count":1,"area":"alignment"},{"id":"2608.13482","title":"Synthetic Persona Pretraining: Alignment from Token Zero","created":"2026-08-13","cited_by_count":1,"area":null},{"id":"2510.05534","title":"Revisiting Self-Play Preference Optimization: On the Role of Prompt Difficulty","created":"2025-10-07","cited_by_count":0,"area":"alignment"},{"id":"2601.14007","title":"BACH-V: Bridging Abstract and Concrete Human-Values in Large Language Models","created":"2026-01-20","cited_by_count":0,"area":null},{"id":"2604.17691","title":"SafeAnchor: Preventing Cumulative Safety Erosion in Continual Domain Adaptation of Large Language Models","created":"2026-04-20","cited_by_count":0,"area":"alignment"},{"id":"2605.02897","title":"Same Voice, Different Lab: On the Homogenization of Frontier LLM Personalities","created":"2026-03-20","cited_by_count":0,"area":null},{"id":"2605.02958","title":"Tracing the Dynamics of Refusal: Exploiting Latent Refusal Trajectories for Robust Jailbreak Detection","created":"2026-05-02","cited_by_count":0,"area":"safety"},{"id":"2605.04572","title":"From Parameter Dynamics to Risk Scoring : Quantifying Sample-Level Safety Degradation in LLM Fine-tuning","created":"2026-05-06","cited_by_count":0,"area":"alignment"},{"id":"2605.11712","title":"Toward Stable Value Alignment: Introducing Independent Modules for Consistent Value Guidance","created":"2026-05-12","cited_by_count":0,"area":"alignment"},{"id":"2605.24583","title":"Measuring Alignment-Induced Activation Shifts Correctly: A Template-Controlled Difference-in-Differences Protocol","created":"2026-05-23","cited_by_count":0,"area":null},{"id":"2606.02322","title":"Repurposing Adversarial Perturbations for Continual Learning: From Defense to Active Alignment","created":"2026-06-01","cited_by_count":0,"area":null},{"id":"2606.03648","title":"Safety Measurements for Fine-tuned LLMs Should be Grounded in Capability","created":"2026-06-02","cited_by_count":0,"area":null},{"id":"2606.28525","title":"A Gravitational Interpretation of Safety Reversion under Fine-Tuning","created":"2026-06-26","cited_by_count":0,"area":"safety"},{"id":"2607.27232","title":"Sympathetic Framing: Evaluating AI Alignment across Sociodemographic Groups","created":"2026-07-22","cited_by_count":0,"area":null},{"id":"2608.05158","title":"Safe Evolution with Circuit Anchors","created":"2026-05-24","cited_by_count":0,"area":null},{"id":"2609.35225","title":"SignFLIP: A Unified Model for Sign Language Translation and Generation via Stage-wise Alignment at Scale","created":"2026-09-28","cited_by_count":0,"area":null},{"id":"2610.02839","title":"To Explore The Strange New World Beyond Data Distribution: System Behavior, Causality Tax, and Non-causal Base Model","created":"2026-10-02","cited_by_count":0,"area":null},{"id":"2610.19365","title":"DANGERLENS: Probing Dangerous Representations Hidden by Safety Alignment","created":"2026-10-14","cited_by_count":0,"area":"alignment"}],"next_cursor":null}