English(EN) Base Models Can Reason By Taking a Cue From Training Data

训练数据提示解锁基础模型推理能力，可媲美RL性能

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-05 00:00

研究人员发现，基础模型响应中的特定起始标记可以显著提高其推理能力，使其性能接近通过强化学习（RL）训练的模型。通过操纵这些标记提示，例如使用“\n\nOkay”或“Alright,”，在MATH-500等数学基准测试上的性能可以大幅提升。研究进一步揭示，这些提示是从训练数据中学习到的，而RL训练使它们出现的概率更高。对训练数据进行因果干预甚至可以将任意词语变成有效的推理提示，这表明数据内容与模型行为（包括安全相关响应）之间存在直接联系。 AI

影响表明通过特定数据提示对基础模型进行微调，可以在没有大量RL训练的情况下显著提升推理能力。

排序理由学术论文，详细介绍了关于模型行为的新发现。[lever_c_demoted from research: ic=1 ai=1.0]

在 Hugging Face Daily Papers 阅读 →

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

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 学术论文，详细介绍了关于模型行为的新发现。[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: 3 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

Hugging Face Daily Papers TIER_1 English(EN) · 2026-10-05 00:00

基础模型可通过借鉴训练数据进行推理

In this paper, we study how training data creates associations between the tokens at the start of a base model's response and the reasoning behavior that follows. First, we demonstrate that fixing particular starting token cues makes a base model's performance competitive with th…

本文如何被排名

报道来源 [1]

基础模型可通过借鉴训练数据进行推理

相关实体

相关话题