{"id":"arxiv/2609.21960","url":"https://arxiv.org/abs/2609.21960","tags":["math.ST","cs.LG","stat.ML"],"score":0,"title":"Schedule optimization for tau-leaping in masked discrete diffusion","domain":"math.ST","source":"arxiv","authors":["Cecilia Secchi","Giacomo Zanella"],"reviews":0,"subject":"math.ST","summary":"Masked diffusions are popular generative models for discrete distributions. Unlike standard autoregressive sampling, they reveal several coordinates in parallel, approximating each block's joint conditional law by a product of one-coordinate conditionals. The resulting procedure, usually called tau-leaping, reduces computational cost but introduces a factorization error ($\\varepsilon_\\text{fact}$), even with perfectly learned predictors. We study the resulting tradeoff between generative accuracy and computational cost, focusing on how to choose a denoising schedule to minimize…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-09-18","reactions":{}}