Beyond In-Distribution Success: Scaling Curves of CoT Granularity for Language Model Generalization

Ru Wang, Wei Huang, Selena Song, Haoyu Zhang, Qian Niu, Yusuke Iwasawa, Yutaka Matsuo, Jiaxian Guo

Keywords: Chain of Thought, Scaling Curve, Out-of-Distribution Generalization, Sample Efficiency

TL;DR: CoT supervision enables transformers to generalize under distribution shifts, outperforming QA training with higher OOD accuracy, finer granularity, and up to 80% fewer examples, by enforcing valid reasoning structures amplified by recap conditions

Abstract: Generalization to novel compound tasks under distribution shift is important for deploying transformer-based language models (LMs). This work investigates Chain-of-Thought (CoT) reasoning as a means to enhance OOD generalization. Through controlled experiments across several compound tasks, we reveal three key insights: (1) While QA-trained models achieve near-perfect in-distribution accuracy, their OOD performance degrades catastrophically, even with 10000k+ training examples; (2) the granularity of CoT data strongly correlates with generalization performance; finer-grained CoT data leads to better generalization; (3) CoT exhibits remarkable sample efficiency, matching QA performance with much less (even 80%) data. Theoretically, we demonstrate that CoT forces internalization of valid dependency structures, and thus can achieve better generalization. Further, we show that transformer positional embeddings can amplify generalization by emphasizing subtask condition recurrence in long CoT sequences. Our combined theoretical and empirical analysis provides compelling evidence for CoT reasoning as a crucial training paradigm for enabling LM generalization on multi-step reasoning tasks under structural distributional shifts.

Submission Number: 119