Steering Away from Memorization: Reachability-Constrained Reinforcement Learning for Text-to-Image Diffusion
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866912931785998336 |
|---|---|
| author | Karnik, Sathwik Kim, Juyeop Koyejo, Sanmi Lee, Jong-Seok Bansal, Somil |
| author_facet | Karnik, Sathwik Kim, Juyeop Koyejo, Sanmi Lee, Jong-Seok Bansal, Somil |
| contents | Text-to-image diffusion models often memorize training data, revealing a fundamental failure to generalize beyond the training set. Current mitigation strategies typically sacrifice image quality or prompt alignment to reduce memorization. To address this, we propose Reachability-Aware Diffusion Steering (RADS), an inference-time framework that prevents memorization while preserving generation fidelity. RADS models the diffusion denoising process as a dynamical system and applies concepts from reachability analysis to approximate the "backward reachable tube"--the set of intermediate states that inevitably evolve into memorized samples. We then formulate mitigation as a constrained reinforcement learning (RL) problem, where a policy learns to steer the trajectory away from memorization via minimal perturbations in the caption embedding space. Empirical evaluations show that RADS achieves a superior Pareto frontier between generation diversity (SSCD), quality (FID), and alignment (CLIP) compared to state-of-the-art baselines. Crucially, RADS provides robust mitigation without modifying the diffusion backbone, offering a plug-and-play solution for safe generation. Our website is available at: https://s-karnik.github.io/rads-memorization-project-page/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_00140 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Steering Away from Memorization: Reachability-Constrained Reinforcement Learning for Text-to-Image Diffusion Karnik, Sathwik Kim, Juyeop Koyejo, Sanmi Lee, Jong-Seok Bansal, Somil Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Text-to-image diffusion models often memorize training data, revealing a fundamental failure to generalize beyond the training set. Current mitigation strategies typically sacrifice image quality or prompt alignment to reduce memorization. To address this, we propose Reachability-Aware Diffusion Steering (RADS), an inference-time framework that prevents memorization while preserving generation fidelity. RADS models the diffusion denoising process as a dynamical system and applies concepts from reachability analysis to approximate the "backward reachable tube"--the set of intermediate states that inevitably evolve into memorized samples. We then formulate mitigation as a constrained reinforcement learning (RL) problem, where a policy learns to steer the trajectory away from memorization via minimal perturbations in the caption embedding space. Empirical evaluations show that RADS achieves a superior Pareto frontier between generation diversity (SSCD), quality (FID), and alignment (CLIP) compared to state-of-the-art baselines. Crucially, RADS provides robust mitigation without modifying the diffusion backbone, offering a plug-and-play solution for safe generation. Our website is available at: https://s-karnik.github.io/rads-memorization-project-page/. |
| title | Steering Away from Memorization: Reachability-Constrained Reinforcement Learning for Text-to-Image Diffusion |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2603.00140 |