{
  "id": "2611.06108",
  "title": "Learning When to Think Aloud in Latent Reasoning",
  "authors": "Zhehao Zhou, Tongrui Su, Shengyu Zhu, Da Li, Zeyu Liu, Jun Zhu, Xueqi Cheng",
  "first_public": "2026-11-04",
  "added_at": "2026-11-09",
  "venue": "arXiv",
  "arxiv": "2611.06108",
  "paper_url": "https://arxiv.org/abs/2611.06108",
  "html_url": "https://arxiv.org/html/2611.06108v1",
  "lens": [
    "continuous"
  ],
  "representation": [
    "text-latent-hybrid",
    "continuous-state"
  ],
  "computation": [
    "token-wise"
  ],
  "training": [
    "curriculum",
    "supervised",
    "reinforcement-learning"
  ],
  "questions": [
    "efficiency",
    "controllability",
    "capability"
  ],
  "evidence": [
    "empirical"
  ],
  "modality": [
    "language"
  ],
  "summary": "Learns when to write an intermediate token and when to feed a hidden state back into the model. Retains important words from each reasoning step as training anchors, then trains the mixed text-and-state policy with answer and length rewards.",
  "provenance": "Primary-source literature review, 2026-11-09",
  "author_names": [
    "Zhehao Zhou",
    "Tongrui Su",
    "Shengyu Zhu",
    "Da Li",
    "Zeyu Liu",
    "Jun Zhu",
    "Xueqi Cheng"
  ],
  "record_url": "https://latent-implicit-thinking.github.io/papers/2611.06108/",
  "legacy_record_url": "https://latent-implicit-thinking.github.io/papers/#paper-2611-06108",
  "citation_date": "2026-11-04",
  "json_url": "https://latent-implicit-thinking.github.io/papers/2611.06108/record.json",
  "markdown_url": "https://latent-implicit-thinking.github.io/papers/2611.06108/index.md",
  "bibtex_url": "https://latent-implicit-thinking.github.io/papers/2611.06108/citation.bib",
  "csl_url": "https://latent-implicit-thinking.github.io/papers/2611.06108/citation.csl.json",
  "pdf_url": "https://arxiv.org/pdf/2611.06108"
}