Cover Image

GR-SAP: Generative Replay for Safety Alignment Preservation during Fine-Tuning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Fang, Zhouxiang, Zhou, Jiawei, Chen, Hanjie
Format:	Preprint
Published:	2026
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Benchmarking Large Language Models on Answering and Explaining Challenging Medical Questions
by: Chen, Hanjie, et al.
Published: (2024)

PRISM: A Dual View of LLM Reasoning through Semantic Flow and Latent Computation
by: Chang, Ruidi, et al.
Published: (2026)

Few Tokens, Big Leverage: Preserving Safety Alignment by Constraining Safety Tokens during Fine-tuning
by: Wang, Guoli, et al.
Published: (2026)

DEFT: Distribution-guided Efficient Fine-Tuning for Human Alignment
by: Zhu, Liang, et al.
Published: (2026)

MSSR: Memory-Aware Adaptive Replay for Continual LLM Fine-Tuning
by: Lu, Yiyang, et al.
Published: (2026)

SafeMERGE: Preserving Safety Alignment in Fine-Tuned Large Language Models via Selective Layer-Wise Model Merging
by: Djuhera, Aladin, et al.
Published: (2025)

RefusalGuard: Geometry-Preserving Fine-Tuning for Safety in LLMs
by: Asif, Sadia, et al.
Published: (2026)

NeuronTune: Fine-Grained Neuron Modulation for Balanced Safety-Utility Alignment in LLMs
by: Pan, Birong, et al.
Published: (2025)

The Generalization Ridge: Information Flow in Natural Language Generation
by: Chang, Ruidi, et al.
Published: (2025)

Be Your Own Red Teamer: Safety Alignment via Self-Play and Reflective Experience Replay
by: Wang, Hao, et al.
Published: (2026)

Re-Emergent Misalignment: How Narrow Fine-Tuning Erodes Safety Alignment in LLMs
by: Giordani, Jeremiah
Published: (2025)

Separate the Wheat from the Chaff: A Post-Hoc Approach to Safety Re-Alignment for Fine-Tuned Language Models
by: Wu, Di, et al.
Published: (2024)

Stealth Fine-Tuning: Efficiently Breaking Alignment in RVLMs Using Self-Generated CoT
by: Yu, Le, et al.
Published: (2025)

Fine-Tuning Lowers Safety and Disrupts Evaluation Consistency
by: Fraser, Kathleen C., et al.
Published: (2025)

Decoupling Safety into Orthogonal Subspace: Cost-Efficient and Performance-Preserving Alignment for Large Language Models
by: Mou, Yutao, et al.
Published: (2025)

Unforgotten Safety: Preserving Safety Alignment of Large Language Models with Continual Learning
by: Alssum, Lama, et al.
Published: (2025)

Intuitive Fine-Tuning: Towards Simplifying Alignment into a Single Process
by: Hua, Ermo, et al.
Published: (2024)

SPP: Sparsity-Preserved Parameter-Efficient Fine-Tuning for Large Language Models
by: Lu, Xudong, et al.
Published: (2024)

ICL CIPHERS: Quantifying "Learning" in In-Context Learning via Substitution Ciphers
by: Fang, Zhouxiang, et al.
Published: (2025)

Safety-Aware Fine-Tuning of Large Language Models
by: Choi, Hyeong Kyu, et al.
Published: (2024)

SAFENLIDB: A Privacy-Preserving Safety Alignment Framework for LLM-based Natural Language Database Interfaces
by: Liu, Ruiheng, et al.
Published: (2025)

Toward Secure Tuning: Mitigating Security Risks from Instruction Fine-Tuning
by: Du, Yanrui, et al.
Published: (2024)

Why Reinforcement Fine-Tuning Enables MLLMs Preserve Prior Knowledge Better: A Data Perspective
by: Zhang, Zhihao, et al.
Published: (2025)

Learning to Stay Safe: Adaptive Regularization Against Safety Degradation during Fine-Tuning
by: Goel, Jyotin, et al.
Published: (2026)

Mitigating Fine-tuning based Jailbreak Attack with Backdoor Enhanced Safety Alignment
by: Wang, Jiongxiao, et al.
Published: (2024)

SIBO: A Simple Booster for Parameter-Efficient Fine-Tuning
by: Wen, Zhihao, et al.
Published: (2024)

Long Is More for Alignment: A Simple but Tough-to-Beat Baseline for Instruction Fine-Tuning
by: Zhao, Hao, et al.
Published: (2024)

Scaling Data Diversity for Fine-Tuning Language Models in Human Alignment
by: Song, Feifan, et al.
Published: (2024)

Learning Distribution-Wise Control in Representation Space for Language Models
by: Deng, Chunyuan, et al.
Published: (2025)

Steering Information Utility in Key-Value Memory for Language Model Post-Training
by: Deng, Chunyuan, et al.
Published: (2025)

Cause-Aware Empathetic Response Generation via Chain-of-Thought Fine-Tuning
by: Chen, Xinhao, et al.
Published: (2024)

Privacy-Preserving Parameter-Efficient Fine-Tuning for Large Language Model Services
by: Li, Yansong, et al.
Published: (2023)

COPU: Conformal Prediction for Uncertainty Quantification in Natural Language Generation
by: Wang, Sean, et al.
Published: (2025)

Rethinking Bottlenecks in Safety Fine-Tuning of Vision Language Models
by: Ding, Yi, et al.
Published: (2025)

GPTA: Generative Prompt Tuning Assistant for Synergistic Downstream Neural Network Enhancement with LLMs
by: Liu, Xiao, et al.
Published: (2024)

Fine-Tuning Language Models for Ethical Ambiguity: A Comparative Study of Alignment with Human Responses
by: Senthilkumar, Pranav, et al.
Published: (2024)

STAIR: Improving Safety Alignment with Introspective Reasoning
by: Zhang, Yichi, et al.
Published: (2025)

Controlling Distributional Bias in Multi-Round LLM Generation via KL-Optimized Fine-Tuning
by: Jiang, Yanbei, et al.
Published: (2026)

Fine-Tuning on Noisy Instructions: Effects on Generalization and Performance
by: Alajrami, Ahmed, et al.
Published: (2025)

Supervised In-Context Fine-Tuning for Generative Sequence Labeling
by: Dukić, David, et al.
Published: (2025)