הכוונה של דיפוזיה ספציפית-מודעה באמצעות סטטיסטיקה מונטה-קרלו מורכבת

תקציר מקורי באנגליתarXiv:2610.00395v1 Announce Type: new Abstract: Inference-time steering enables pretrained diffusion models to satisfy new constraints without full retraining. However, specificity-aware generation is difficult: repelling samples from a negative reference distribution can also erode the positive distribution where the two overlap. The key challenge is to suppress negative mass while minimally distorting the positive distribution. We address this problem by formulating specificity-aware steering as a target-design problem and deriving a target distribution from an overlap-based objective. The resulting target keeps the desired reference distribution only in regions where it is sufficiently preferred over the undesired reference distribution, giving a likelihood-ratio interpretation of speci