English

Specializing Smaller Language Models towards Multi-Step Reasoning

Computation and Language 2023-01-31 v1 Artificial Intelligence Machine Learning

Abstract

The surprising ability of Large Language Models (LLMs) to perform well on complex reasoning with only few-shot chain-of-thought prompts is believed to emerge only in very large-scale models (100+ billion parameters). We show that such abilities can, in fact, be distilled down from GPT-3.5 (\ge 175B) to T5 variants (\le 11B). We propose model specialization, to specialize the model's ability towards a target task. The hypothesis is that large models (commonly viewed as larger than 100B) have strong modeling power, but are spread on a large spectrum of tasks. Small models (commonly viewed as smaller than 10B) have limited model capacity, but if we concentrate their capacity on a specific target task, the model can achieve a decent improved performance. We use multi-step math reasoning as our testbed because it is a very typical emergent ability. We show two important aspects of model abilities: (1). there exists a very complex balance/ tradeoff between language models' multi-dimensional abilities; (2). by paying the price of decreased generic ability, we can clearly lift up the scaling curve of models smaller than 10B towards a specialized multi-step math reasoning ability. We further give comprehensive discussions about important design choices for better generalization, including the tuning data format, the start model checkpoint, and a new model selection method. We hope our practice and discoveries can serve as an important attempt towards specialized smaller models in the new research paradigm set by LLMs.

Keywords

Cite

@article{arxiv.2301.12726,
  title  = {Specializing Smaller Language Models towards Multi-Step Reasoning},
  author = {Yao Fu and Hao Peng and Litu Ou and Ashish Sabharwal and Tushar Khot},
  journal= {arXiv preprint arXiv:2301.12726},
  year   = {2023}
}

Comments

Preprint

R2 v1 2026-06-28T08:26:10.268Z