SoT: Structured-of-Thought Prompting Guides Multilingual Reasoning in Large Language Models
This addresses the challenge of enabling complex reasoning in non-high-resource languages for AI applications, representing an incremental improvement through a novel prompting technique.
The paper tackles the problem of poor multilingual reasoning in large language models due to resource constraints by proposing Structured-of-Thought (SoT), a training-free method that improves performance on multilingual reasoning benchmarks, outperforming strong baselines across various LLM backbones.
Recent developments have enabled Large Language Models (LLMs) to engage in complex reasoning tasks through deep thinking. However, the capacity of reasoning has not been successfully transferred to non-high-resource languages due to resource constraints, which struggles with multilingual reasoning tasks. To this end, we propose Structured-of-Thought (SoT), a training-free method that improves the performance on multilingual reasoning through a multi-step transformation: Language Thinking Transformation and Structured Knowledge Transformation. The SoT method converts language-specific semantic information into language-agnostic structured representations, enabling the models to understand the query in different languages more sophisticated. Besides, SoT effectively guides LLMs toward more concentrated reasoning to maintain consistent underlying reasoning pathways when handling cross-lingual variations in expression. Experimental results demonstrate that SoT outperforms several strong baselines on multiple multilingual reasoning benchmarks when adapting to various backbones of LLMs. It can also be integrated with other training-free strategies for further improvements. Our code is available at https://github.com/Cherry-qwq/SoT.