{"total":64,"rows":[{"id":"2509.06948","title":"Beyond Two-Stage Training: Cooperative SFT and RL for LLM Reasoning","created":"2025-09-08","cited_by_count":30,"area":"lr"},{"id":"2507.18631","title":"Layer-Aware Representation Filtering: Purifying Finetuning Data to Preserve LLM Safety Alignment","created":"2025-07-24","cited_by_count":24,"area":"alignment"},{"id":"2502.20968","title":"Beware of Your Po! Measuring and Mitigating AI Safety Risks in Role-Play Fine-Tuning of LLMs","created":"2025-02-28","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":"2501.18100","title":"Panacea: Mitigating Harmful Fine-tuning for Large Language Models via Post-fine-tuning Perturbation","created":"2025-01-30","cited_by_count":20,"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":"2506.08473","title":"AsFT: Anchoring Safety During LLM Fine-Tuning Within Narrow Safety Basin","created":"2025-06-10","cited_by_count":19,"area":null},{"id":"2503.20807","title":"Fundamental Safety-Capability Trade-offs in Fine-tuning Large Language Models","created":"2025-03-24","cited_by_count":17,"area":null},{"id":"2505.17895","title":"DataRater: Meta-Learned Dataset Curation","created":"2025-05-23","cited_by_count":13,"area":null},{"id":"2510.27172","title":"Adaptive Defense against Harmful Fine-Tuning for Large Language Models via Bayesian Data Scheduler","created":"2025-10-31","cited_by_count":12,"area":null},{"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":"2505.17196","title":"Shape it Up! Restoring LLM Safety during Finetuning","created":"2025-05-22","cited_by_count":11,"area":"alignment"},{"id":"2508.07172","title":"Gradient Surgery for Safe LLM Fine-Tuning","created":"2025-08-10","cited_by_count":11,"area":"alignment"},{"id":"2510.14301","title":"A Guardrail for Safety Preservation: When Safety-Sensitive Subspace Meets Harmful-Resistant Null-Space","created":"2025-10-16","cited_by_count":9,"area":"alignment"},{"id":"2509.22745","title":"Defending MoE LLMs against Harmful Fine-Tuning via Safety Routing Alignment","created":"2025-09-26","cited_by_count":8,"area":"moe"},{"id":"2601.01887","title":"Safety at One Shot: Patching Fine-Tuned LLMs with A Single Instance","created":"2026-01-05","cited_by_count":8,"area":"alignment"},{"id":"2502.00840","title":"Activation Approximations Can Incur Safety Vulnerabilities Even in Aligned LLMs: Comprehensive Analysis and Defense","created":"2025-02-02","cited_by_count":7,"area":null},{"id":"2502.08585","title":"LDC-MTL: Balancing Multi-Task Learning through Scalable Loss Discrepancy Control","created":"2025-02-12","cited_by_count":7,"area":null},{"id":"2502.19537","title":"No, of Course I Can! Deeper Fine-Tuning Attacks That Bypass Token-Level Safety Mechanisms","created":"2025-02-26","cited_by_count":6,"area":"safety"},{"id":"2505.14185","title":"Safety Subspaces are Not Linearly Distinct: A Fine-Tuning Case Study","created":"2025-05-20","cited_by_count":6,"area":"alignment"},{"id":"2506.07452","title":"When Style Breaks Safety: Defending LLMs Against Superficial Style Alignment","created":"2025-06-09","cited_by_count":6,"area":"safety"},{"id":"2507.17075","title":"LoRA is All You Need for Safety Alignment of Reasoning LLMs","created":"2025-07-22","cited_by_count":6,"area":"lr"},{"id":"2510.04320","title":"Read the Scene, Not the Script: Outcome-Aware Safety for LLMs","created":"2025-10-05","cited_by_count":6,"area":"alignment"},{"id":"2510.10085","title":"Pharmacist: Safety Alignment Data Curation for Large Language Models against Harmful Fine-tuning","created":"2025-10-11","cited_by_count":6,"area":"alignment"},{"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":"2502.08808","title":"A First-order Generative Bilevel Optimization Framework for Diffusion Models","created":"2025-02-12","cited_by_count":4,"area":"diffusion"},{"id":"2505.18458","title":"A Survey of LLM $\\times$ DATA","created":"2025-05-24","cited_by_count":4,"area":"rag"},{"id":"2602.05228","title":"Surgery: Mitigating Harmful Fine-Tuning for Large Language Models via Attention Sink","created":"2026-02-05","cited_by_count":4,"area":"alignment"},{"id":"2510.06048","title":"BLISS: A Lightweight Bilevel Influence Scoring Method for Data Selection in Language Model Pretraining","created":"2025-10-07","cited_by_count":3,"area":null},{"id":"2601.08000","title":"Reasoning over Precedents Alongside Statutes: Case-Augmented Deliberative Alignment for LLM Safety","created":"2026-01-12","cited_by_count":3,"area":"lr"},{"id":"2603.01185","title":"Token-level Data Selection for Safe LLM Fine-tuning","created":"2026-03-01","cited_by_count":3,"area":null},{"id":"2507.07723","title":"Stable Preference Optimization: A Bilevel Approach to Catastrophic Preference Shift","created":"2025-07-10","cited_by_count":2,"area":"alignment"},{"id":"2511.21056","title":"Bilevel Data Curation for LLM Fine-tuning: Offline Selection and Online Self-Refining Generation","created":"2025-11-26","cited_by_count":2,"area":null},{"id":"2601.07200","title":"Safeguarding LLM Fine-tuning via Push-Pull Distributional Alignment","created":"2026-01-12","cited_by_count":2,"area":"alignment"},{"id":"2602.16835","title":"NeST: Neuron Selective Tuning for LLM Safety","created":"2026-02-18","cited_by_count":2,"area":"alignment"},{"id":"2506.00676","title":"SafeTuneBed: A Toolkit for Benchmarking LLM Safety Alignment in Fine-Tuning","created":"2025-05-31","cited_by_count":1,"area":"alignment"},{"id":"2509.02937","title":"Faster Gradient Methods for Highly-Smooth Stochastic Bilevel Optimization","created":"2025-09-03","cited_by_count":1,"area":null},{"id":"2509.15543","title":"Nonconvex Decentralized Stochastic Bilevel Optimization under Heavy-Tailed Noise","created":"2025-09-19","cited_by_count":1,"area":null},{"id":"2510.14459","title":"Holdout-Loss-Based Data Selection for LLM Finetuning via In-Context Learning","created":"2025-10-16","cited_by_count":1,"area":"alignment"},{"id":"2510.21885","title":"Preventing Catastrophic Forgetting: Behavior-Aware Sampling for Safer Language Model Fine-Tuning","created":"2025-10-23","cited_by_count":1,"area":null},{"id":"2512.16905","title":"Alchemist: Unlocking Efficiency in Text-to-Image Model Training via Meta-Gradient Data Selection","created":"2025-12-18","cited_by_count":1,"area":"diffusion"},{"id":"2602.04306","title":"DeFrame: Debiasing Large Language Models Against Framing Effects","created":"2026-02-04","cited_by_count":1,"area":null},{"id":"2603.24857","title":"AI Security in the Foundation Model Era: A Comprehensive Survey from a Unified Perspective","created":"2026-03-25","cited_by_count":1,"area":"safety"},{"id":"2605.09094","title":"A Tale of Two Problems: Multi-Task Bilevel Learning Meets Equality Constrained Multi-Objective Optimization","created":"2026-05-09","cited_by_count":1,"area":null},{"id":"2605.25073","title":"Security in the Fine-Tuning Lifecycle of Large Language Models: Threats, Defenses,Evaluation, and Future Directions","created":"2026-05-24","cited_by_count":1,"area":"safety"},{"id":"2605.29396","title":"Aligned but Fragile: Enhancing LLM Safety Robustness via Zeroth-Order Optimization","created":"2026-05-28","cited_by_count":1,"area":"alignment"},{"id":"2606.00160","title":"DataShield: Safety-degrading Data Filtering for LLM Benign Instruction Fine-Tuning","created":"2026-05-29","cited_by_count":1,"area":null},{"id":"2505.17151","title":"Bayesian Optimization for Enhanced Language Models: Optimizing Acquisition Functions","created":"2025-05-22","cited_by_count":0,"area":null},{"id":"2511.09880","title":"EnchTable: Unified Safety Alignment Transfer in Fine-tuned Large Language Models","created":"2025-11-13","cited_by_count":0,"area":"alignment"},{"id":"2601.22016","title":"TBDFiltering: Sample-Efficient Tree-Based Data Filtering","created":"2026-01-29","cited_by_count":0,"area":null},{"id":"2605.06166","title":"One Algorithm, Two Goals: Dual Scoring for Parameter and Data Selection in LLM Fine-Tuning","created":"2026-05-07","cited_by_count":0,"area":null},{"id":"2605.10288","title":"BROS: Bias-Corrected Randomized Subspaces for Memory-Efficient Single-Loop Bilevel Optimization","created":"2026-05-11","cited_by_count":0,"area":null},{"id":"2605.14194","title":"GradShield: Alignment Preserving Finetuning","created":"2026-05-13","cited_by_count":0,"area":null},{"id":"2605.28664","title":"Activation Steering for Synthetic Data Generation: The Role of Diversity in Downstream Safety Detection","created":"2026-05-27","cited_by_count":0,"area":"mechinterp"},{"id":"2607.10263","title":"Sharper Analysis of Single-Loop Methods for Bilevel Optimization","created":"2026-07-11","cited_by_count":0,"area":"deeprl"},{"id":"2608.04322","title":"DataRx: Missingness-Aware Sampling for Safer Large Language Model Task-Specific Fine-Tuning","created":"2026-08-05","cited_by_count":0,"area":null},{"id":"2608.28361","title":"GRACE:Gradient-guided Coreset Selection for LLM Unlearning","created":"2026-08-28","cited_by_count":0,"area":"safety"},{"id":"2609.16694","title":"Toward Secure AI-Powered Penetration Testing Agents: Security Threats, Guardrails, and Architectural Perspectives","created":"2026-09-15","cited_by_count":0,"area":"agents"},{"id":"2609.22144","title":"Multilingual Safety Signals Are Multi-Layered: Filtering Safety-Degrading Data for Safer LLMs","created":"2026-08-26","cited_by_count":0,"area":"alignment"},{"id":"2609.22153","title":"SafeTune: A Unified Faithful Library for Auditing and Repairing Safety Drift in Fine-Tuned LLMs","created":"2026-08-26","cited_by_count":0,"area":null},{"id":"2609.30501","title":"To Solve Bilevel Optimization with Nonconvex Lower Levels, We Need Second-Order Stationarity","created":"2026-09-24","cited_by_count":0,"area":null},{"id":"2609.32484","title":"Towards Scalable Data Diversification for Language Model Pretraining via Leverage Score Sampling","created":"2026-09-26","cited_by_count":0,"area":null},{"id":"2609.36505","title":"BRIDGE: Bilevel Retrieval-Credit-Aware Agentic Reinforcement Learning","created":"2026-09-29","cited_by_count":0,"area":"agents"},{"id":"2610.02092","title":"Scalable, Transferable Meta-network for Data Selection Requires a Different Loss (and Why the Obvious Choice is Problematic)","created":"2026-10-01","cited_by_count":0,"area":null}],"next_cursor":null}