Reasoning Like Program Executors

By Xinyu Pi, Qian Liu, Morteza Ziyadi, Zeqi Lin, Yan Gao, Qiang Fu, Jian-Guang Lou, Weizhu Chen. 2022

61 citations recorded by the source corpus.

Reasoning over natural language is a long-standing goal for the research community. However, studies have shown that existing language models are inadequate in reasoning. To address the issue, we present POET, a novel reasoning pre-training paradigm. Through pre-training language models with programs and their execution results, POET empowers language models to harvest the reasoning knowledge possessed by program executors via a data-driven approach. POET is conceptually simple and can be instantiated by different kinds of program executors. In this paper, we showcase two simple instances POET-Math and POET-Logic, in addition to a complex instance, POET-SQL. Experimental results on six benchmarks demonstrate that POET can significantly boost model performance in natural language reasoning, such as numerical reasoning, logical reasoning, and multi-hop reasoning. POET opens a new gate on reasoning-enhancement pre-training, and we hope our analysis would shed light on the future research of reasoning like program executors.

Subfields: Prompting & Reasoning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Prompting & Reasoning

Citations in the corpus

PAL: Program-aided Language Models (2022, methodology): “Pi et al. (2022) pretrain the model on execution results of random expressions on a calculator, instead of using the solver at test time as well.”
Making Language Models Better Reasoners with Step-Aware Verifier (2022, background): “…Campagna et al. (2020); Wang et al. (2022a) proposed to equip language models with reasoning abilities by generating training examples with human-designed templates; Pi et al. (2022) proposed to inject reasoning capabilities into language models by continual pre-training on program execution data.”
Towards Reasoning in Large Language Models: A Survey (2022, background): “Pi et al. (2022) show that continually pretraining with SQL data can boost the performance of language models, e.”
A Survey of Large Language Models (2023, unknown): “To tackle this issue, a direct way is to tune LLMs on synthesized arithmetic problems [376].”