A LoRA is Worth a Thousand Pictures

arXiv:2412.12048v2 Announce Type: replace Abstract: Low-Rank Adaptation (LoRA) has become a standard approach for parameter-efficient fine-tuning of image diffusion models, with over 200K models trained to replicate artistic styles and shared in online communities. One way to analyze styles of LoRA models is to generate image samples and apply image-based features like CLIP. However, this generation requires additional computation and prompt selection. In this work, we propose analyzing styles directly in the LoRA weight space, bypassing these limitations while achieving better performance even on unseen LoRAs. Since in-the-wild LoRAs lack ground truth model details, we design an experimental setup and show that LoRA weights alone serve as effective feature representations of adapted styles, robust to variations in fine-tuning seeds and image selection. We extensively evaluate using SD1.5 (the most popular base model within online communities) across data domains (fine-art paintings, architectural photos) and diverse prompt settings, with a small-scale SDXL experiment further verifying generalization. These evaluations confirm that our approach, under the selected aspects of variability, preserves fine-grained styles and outperforms image-based methods in clustering and retrieval of unseen LoRAs.

Related

• InstructMoLE: Instruction-Guided Mixture of Low-rank Experts for Multi-Conditional Image Generation
• ACE-LoRA: Adaptive Orthogonal Decoupling for Continual Image Editing
• LoRA on the Go: Instance-level Dynamic LoRA Selection and Merging
• A Fine-Grained Analysis of the LoRA Fine-Tuning Landscape with Implications for Data Selection

Source: arXiv cs.CV | 2026-10-15