Why Do Multilingual Reasoning Gaps Emerge in Reasoning Language Models?

TL;DR

This paper identifies language understanding failures as the main cause of multilingual reasoning gaps in reasoning language models and proposes a selective translation method to mitigate this issue effectively.

Contribution

It demonstrates that understanding failures can be detected and mitigated through selective translation, significantly reducing the multilingual reasoning gap in models.

Findings

Understanding failures are detectable with supervised methods.

Selective translation reduces translation needs to about 20%.

The approach nearly closes the multilingual reasoning gap.

Abstract

Reasoning language models (RLMs) achieve strong performance on complex reasoning tasks, yet they still exhibit a multilingual reasoning gap, performing better in high-resource languages than in low-resource ones. While recent efforts have been made to address this gap, its underlying causes remain largely unexplored. In this work, we show that this gap primarily stems from failures in language understanding-specifically, the model's inability to translate multilingual inputs into the language dominating its reasoning traces (typically English). As identifying understanding failures can enable targeted mitigation of the gap, we evaluate a range of detection methods and find that understanding failures are detectable to a meaningful extent, with supervised approaches performing best. Building on this, we propose Selective Translation, a strategy that incorporates an English translation into the initial reasoning trace only when an understanding failure is detected. Experimental results using Qwen3-4B show that Selective Translation substantially bridges the multilingual reasoning gap, achieving near full-translation performance while translating only about 20% of inputs. Together, our results show that failures in language understanding are the primary driver of the multilingual reasoning gap and can be detected and selectively mitigated, clarifying its origin and suggesting a path toward more equitable multilingual reasoning. Our code and data are publicly available at https://github.com/deokhk/RLM_analysis