On Mechanistic Knowledge Localization in Text-to-Image Generative Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909194792206336 |
|---|---|
| author | Basu, Samyadeep Rezaei, Keivan Kattakinda, Priyatham Rossi, Ryan Zhao, Cherry Morariu, Vlad Manjunatha, Varun Feizi, Soheil |
| author_facet | Basu, Samyadeep Rezaei, Keivan Kattakinda, Priyatham Rossi, Ryan Zhao, Cherry Morariu, Vlad Manjunatha, Varun Feizi, Soheil |
| contents | Identifying layers within text-to-image models which control visual attributes can facilitate efficient model editing through closed-form updates. Recent work, leveraging causal tracing show that early Stable-Diffusion variants confine knowledge primarily to the first layer of the CLIP text-encoder, while it diffuses throughout the UNet.Extending this framework, we observe that for recent models (e.g., SD-XL, DeepFloyd), causal tracing fails in pinpointing localized knowledge, highlighting challenges in model editing. To address this issue, we introduce the concept of Mechanistic Localization in text-to-image models, where knowledge about various visual attributes (e.g., "style", "objects", "facts") can be mechanistically localized to a small fraction of layers in the UNet, thus facilitating efficient model editing. We localize knowledge using our method LocoGen which measures the direct effect of intermediate layers to output generation by performing interventions in the cross-attention layers of the UNet. We then employ LocoEdit, a fast closed-form editing method across popular open-source text-to-image models (including the latest SD-XL)and explore the possibilities of neuron-level model editing. Using Mechanistic Localization, our work offers a better view of successes and failures in localization-based text-to-image model editing. Code will be available at https://github.com/samyadeepbasu/LocoGen. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_01008 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | On Mechanistic Knowledge Localization in Text-to-Image Generative Models Basu, Samyadeep Rezaei, Keivan Kattakinda, Priyatham Rossi, Ryan Zhao, Cherry Morariu, Vlad Manjunatha, Varun Feizi, Soheil Computer Vision and Pattern Recognition Identifying layers within text-to-image models which control visual attributes can facilitate efficient model editing through closed-form updates. Recent work, leveraging causal tracing show that early Stable-Diffusion variants confine knowledge primarily to the first layer of the CLIP text-encoder, while it diffuses throughout the UNet.Extending this framework, we observe that for recent models (e.g., SD-XL, DeepFloyd), causal tracing fails in pinpointing localized knowledge, highlighting challenges in model editing. To address this issue, we introduce the concept of Mechanistic Localization in text-to-image models, where knowledge about various visual attributes (e.g., "style", "objects", "facts") can be mechanistically localized to a small fraction of layers in the UNet, thus facilitating efficient model editing. We localize knowledge using our method LocoGen which measures the direct effect of intermediate layers to output generation by performing interventions in the cross-attention layers of the UNet. We then employ LocoEdit, a fast closed-form editing method across popular open-source text-to-image models (including the latest SD-XL)and explore the possibilities of neuron-level model editing. Using Mechanistic Localization, our work offers a better view of successes and failures in localization-based text-to-image model editing. Code will be available at https://github.com/samyadeepbasu/LocoGen. |
| title | On Mechanistic Knowledge Localization in Text-to-Image Generative Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2405.01008 |