Kindly request the inclusion on a line of papers on harmful fine-tuning for LLMs #33
Description
Thank you for the wonderful paper collection. We have a line of research on harmful fine-tuning for LLMs. Could you please include this line of work into the repo?
| Title | Link | Code | Venue | Classification | Model | Comment |
|---|---|---|---|---|---|---|
| Vaccine: Perturbation-aware Alignment for Large Language Models against Harmful Fine-tuning | arxiv | github | NeurIPS'24 | Defense | LLM | Harmful fine-tuning |
| Lazy Safety Alignment for Large Language Models against Harmful Fine-tuning | arxiv | github | NeurIPS'24 | Defense | LLM | Harmful fine-tuning |
| Booster: Tackling Harmful Fine-tuning for Large Language Models via Attenuating Harmful Perturbation | arxiv | github | arXiv | Defense | LLM | Harmful fine-tuning |
| Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning | arxiv | To-be-released | arXiv | Defense | LLM | Harmful fine-tuning |
| Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey | arxiv | awesome project | arXiv | Survey& Other awesome project | LLM | Harmful fine-tuning |
Thank you in advance!
Best,
Tiansheng Huang