{
  "id": 15538077,
  "title": "CoTA++: Understanding the Repeat Curse in dMLLMs from an Information Flow Perspective",
  "url": "https://urgent.news/2026/10/16/cota-understanding-the-repeat-curse-in-dmllms-from-an-information",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-16T10:58:03.000Z",
  "source": {
    "name": "arXiv cs.AI",
    "slug": "arxiv-cs-ai",
    "url": "https://arxiv.org/abs/2610.23290v1"
  },
  "original_language": "en",
  "account": null,
  "summary": "Diffusion-based multimodal large language models (dMLLMs) decode by iterative denoising with bidirectional attention, which precludes the standard autoregressive key--value cache. Practical deployments therefore rely on approximate caching that reuses intermediate states across denoising steps. Such caching, however, often makes a response repeat one token at adjacent positions in a run that…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}