{
  "id": 71,
  "url": "https://arxiv.org/abs/2607.09134v1",
  "title": "ReGen: Hierarchical Multi-Prompt Representation Generation for Efficient Waveform Diffusion Models",
  "summary": "Representation alignment (REPA) has been investigated to accelerate diffusion training, but we observe that regularizing intermediate representations in diffusion Transformers (DiT) may implicitly entangle latents and limit generative capacity. To address this issue, we propose ReGen, a hierarchical multi-prompt representation generation framework that jointly estimates multiple vector fields for both representations and data within a single diffusion model. We further introduce generalized flow",
  "authors": "Sang-Hoon Lee, Ha-Yeong Choi",
  "category": "research",
  "topics": "safety-alignment,finance-investment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-07-10T06:44:04.000Z",
  "fetched_at": "2026-07-14T14:14:15.666Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/71",
  "original_url": "https://arxiv.org/abs/2607.09134v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}