递归语言模型在域外泛化

中文摘要 AI 辅助

我们研究了限制语言模型可见内容何时能改善学习。我们将标准思维链（CoT）——一种读取完整轨迹的更通用的学习器——与递归语言模型进行比较，后者通过在每个子任务的隔离上下文中求解来限制自身。在分布内，这种通用性是免费的：CoT可以高效地模拟递归规则，因此IID泛化保证仅相差一个常数因子，递归并未带来太多优势。但在域外，CoT可能依赖当前子任务之外的上下文来拟合训练，即一种捷径，一旦这些词元改变就会失效；递归的上下文隔离排除了这种失败模式。尽管CoT的类别仍覆盖递归规则，但简单性偏差会选择捷径而非真相。因此，要超越分布精度并真正推理，仅覆盖正确的规则是不够的；这与经典学习理论形成对比。

英文摘要

We study when limiting what a language model can see improves learning. We compare standard CoT, the more general learner that reads the full trace, with recursive language models, which restricts itself by solving each subtask in an isolated context. In-distribution, this generality comes for free: CoT can efficiently simulate the recursive rule, so the IID generalization guarantee changes only by a constant factor, and recursion does not offer much. But out of domain, CoT can fit training by relying on context outside the current subtask, i.e. a shortcut that breaks once those tokens change; recursive context isolation rules out this failure mode. Even though CoT's class still covers the recursive rule, simplicity bias picks the shortcut over the truth. Thus, to go beyond distributional accuracy and truly reason, covering the right rule is not enough; this contrasts with classical learning theory.

发表机构

• Toyota Technological Institute at Chicago（芝加哥丰田理工学院）

机构由 AI 辅助整理，请以论文原文为准。