{"id":"arxiv/2610.01318","url":"https://arxiv.org/abs/2610.01318","tags":["cs.LG","cs.AI","math.ST"],"score":0,"title":"Feature Selective Model Collapse in Diffusion Models: Total Replacement versus Fixed-Budget Training","domain":"cs.LG","source":"arxiv","authors":["Hanna Malet","Gabriel Turinici"],"reviews":0,"subject":"cs.LG","summary":"Model collapse arises when generative models are trained on synthetic data produced by earlier models. The phenomenon has attracted considerable attention because of its societal and technical implications. However, previous studies have reached seemingly contradictory conclusions: replacing real data with synthetic data causes collapse (Shumailov et al.), yet accumulating real data alongside synthetic data can prevent it. For diffusion models, we study an intermediate regime typical of finite-budget pipelines: all past datasets and the real data are kept, but each new model is trained on a…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-10-01","reactions":{}}