{"ID":2868775,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.15865","arxiv_id":"2509.15865","title":"SAGE: Semantic-Aware Shared Sampling for Efficient Diffusion","abstract":"Diffusion models manifest evident benefits across diverse domains, yet their high sampling cost, requiring dozens of sequential model evaluations, remains a major limitation. Prior efforts mainly accelerate sampling via optimized solvers or distillation, which treat each query independently. In contrast, we reduce total number of steps by sharing early-stage sampling across semantically similar queries. To enable such efficiency gains without sacrificing quality, we propose SAGE, a semantic-aware shared sampling framework that integrates a shared sampling scheme for efficiency and a tailored training strategy for quality preservation. Extensive experiments show that SAGE reduces sampling cost by 25.5%, while improving generation quality with 5.0% lower FID, 5.4% higher CLIP, and 160% higher diversity over baselines.","short_abstract":"Diffusion models manifest evident benefits across diverse domains, yet their high sampling cost, requiring dozens of sequential model evaluations, remains a major limitation. Prior efforts mainly accelerate sampling via optimized solvers or distillation, which treat each query independently. In contrast, we reduce tota...","url_abs":"https://arxiv.org/abs/2509.15865","url_pdf":"https://arxiv.org/pdf/2509.15865v1","authors":"[\"Haoran Zhao\",\"Tong Bai\",\"Lei Huang\",\"Xiaoyu Liang\"]","published":"2025-09-19T11:06:52Z","proceeding":"cs.LG","tasks":"[\"cs.LG\"]","methods":"[\"Diffusion Model\"]","has_code":false}