Recovering the Pre-Fine-Tuning Weights of Generative Models
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866911938937618432 |
|---|---|
| author | Horwitz, Eliahu Kahana, Jonathan Hoshen, Yedid |
| author_facet | Horwitz, Eliahu Kahana, Jonathan Hoshen, Yedid |
| contents | The dominant paradigm in generative modeling consists of two steps: i) pre-training on a large-scale but unsafe dataset, ii) aligning the pre-trained model with human values via fine-tuning. This practice is considered safe, as no current method can recover the unsafe, pre-fine-tuning model weights. In this paper, we demonstrate that this assumption is often false. Concretely, we present Spectral DeTuning, a method that can recover the weights of the pre-fine-tuning model using a few low-rank (LoRA) fine-tuned models. In contrast to previous attacks that attempt to recover pre-fine-tuning capabilities, our method aims to recover the exact pre-fine-tuning weights. Our approach exploits this new vulnerability against large-scale models such as a personalized Stable Diffusion and an aligned Mistral. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_10208 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Recovering the Pre-Fine-Tuning Weights of Generative Models Horwitz, Eliahu Kahana, Jonathan Hoshen, Yedid Machine Learning Computation and Language Cryptography and Security Computer Vision and Pattern Recognition The dominant paradigm in generative modeling consists of two steps: i) pre-training on a large-scale but unsafe dataset, ii) aligning the pre-trained model with human values via fine-tuning. This practice is considered safe, as no current method can recover the unsafe, pre-fine-tuning model weights. In this paper, we demonstrate that this assumption is often false. Concretely, we present Spectral DeTuning, a method that can recover the weights of the pre-fine-tuning model using a few low-rank (LoRA) fine-tuned models. In contrast to previous attacks that attempt to recover pre-fine-tuning capabilities, our method aims to recover the exact pre-fine-tuning weights. Our approach exploits this new vulnerability against large-scale models such as a personalized Stable Diffusion and an aligned Mistral. |
| title | Recovering the Pre-Fine-Tuning Weights of Generative Models |
| topic | Machine Learning Computation and Language Cryptography and Security Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2402.10208 |