PulseAugur
实时 06:10:40
English(EN) Wait, Wait, Wait... Why Do Reasoning Models Loop?

研究人员找出 AI 推理模型陷入循环的原因

一篇新研究论文探讨了推理模型(如 DeepSeek-R1)在解决问题时陷入循环的现象。该研究确定了两个主要原因:由于学习难度导致的风险规避,模型倾向于选择更容易的循环动作而非更难的正确动作;以及 Transformer 模型固有的归纳偏倚,倾向于时间相关的错误。虽然增加温度可以减少循环(通过鼓励探索),但它并未解决根本的学习错误,这表明需要进行训练时干预以获得更全面的解决方案。 AI

影响 识别出 Transformer 推理中的核心问题,可能需要新的训练方法来克服。

排序理由 研究论文发布在 arXiv 上,详细介绍了关于 AI 模型行为的发现。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

研究人员找出 AI 推理模型陷入循环的原因

本文如何被排名

Signal score
34 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
研究论文发布在 arXiv 上,详细介绍了关于 AI 模型行为的发现。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准

报道来源 [1]

  1. arXiv cs.LG TIER_1 English(EN) · Charilaos Pipis, Shivam Garg, Vasilis Kontonis, Vaishnavi Shrivastava, Akshay Krishnamurthy, Dimitris Papailiopoulos ·

    等等,等等,等等……为什么推理模型会陷入循环?

    arXiv:2512.12895v2 Announce Type: replace Abstract: Reasoning models (e.g., DeepSeek-R1) generate long chains of thought to solve harder problems, but they often loop, repeating the same text at low temperatures or with greedy decoding. We study why this happens and what role tem…