Abstraction-of-Thought Makes Language Models Better Reasoners
This addresses the limitation of current reasoning methods in language models by improving generalization through abstraction, offering a novel approach for AI reasoning tasks.
The paper tackles the problem of enabling language models to reason with abstraction by introducing Abstraction-of-Thought (AoT), a structured reasoning format that requires varying levels of abstraction, and shows that models finetuned with AoT outperform those using Chain-of-Thought on 23 tasks from Big-Bench Hard.
Abstract reasoning, the ability to reason from the abstract essence of a problem, serves as a key to generalization in human reasoning. However, eliciting language models to perform reasoning with abstraction remains unexplored. This paper seeks to bridge this gap by introducing a novel structured reasoning format called Abstraction-of-Thought (AoT). The uniqueness of AoT lies in its explicit requirement for varying levels of abstraction within the reasoning process. This approach could elicit language models to first contemplate on the abstract level before incorporating concrete details, which is overlooked by the prevailing step-by-step Chain-of-Thought (CoT) method. To align models with the AoT format, we present AoT Collection, a generic finetuning dataset consisting of 348k high-quality samples with AoT reasoning processes, collected via an automated and scalable pipeline. We finetune a wide range of language models with AoT Collection and conduct extensive evaluations on 23 unseen tasks from the challenging benchmark Big-Bench Hard. Experimental results indicate that models aligned to AoT reasoning format substantially outperform those aligned to CoT in many reasoning tasks.