# When Models Reason in Your Language: Controlling Thinking Language Comes at the Cost of Accuracy

arXiv: https://arxiv.org/abs/2505.22888 · HTML: https://mlsanity.com/paper/2505.22888/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 28 of its 46 references that are in the map.
- First posted on arXiv 28 May 2025.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 28 of 28)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://mlsanity.com/paper/2311.12022/) (2023)
- [Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters](https://mlsanity.com/paper/2408.03314/) (2024)
- [s1: Simple test-time scaling](https://mlsanity.com/paper/2501.19393/) (2025)
- [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024)
- [LIMO: Less is More for Reasoning](https://mlsanity.com/paper/2502.03387/) (2025)
- [Reasoning Models Don't Always Say What They Think](https://mlsanity.com/paper/2505.05410/) (2025)
- [Deep Reinforcement Learning: An Overview](https://mlsanity.com/paper/1701.07274/) (2017)
- [Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation](https://mlsanity.com/paper/2412.03304/) (2024)
- [Chain-of-Thought Reasoning In The Wild Is Not Always Faithful](https://mlsanity.com/paper/2503.08679/) (2025)
- [Not All Languages Are Created Equal in LLMs: Improving Multilingual Capability by Cross-Lingual-Thought Prompting](https://mlsanity.com/paper/2305.07004/) (2023)
- [From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models](https://mlsanity.com/paper/2406.16838/) (2024)
- [Are DeepSeek R1 And Other Reasoning Models More Faithful?](https://mlsanity.com/paper/2501.08156/) (2025)
- [Effectively Controlling Reasoning Models through Thinking Intervention](https://mlsanity.com/paper/2503.24370/) (2025)
- [Crosslingual Reasoning through Test-Time Scaling](https://mlsanity.com/paper/2505.05408/) (2025)
- [IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models](https://mlsanity.com/paper/2406.03368/) (2024)
- [Beyond Semantics: The Unreasonable Effectiveness of Reasonless Intermediate Tokens](https://mlsanity.com/paper/2505.13775/) (2025)
- [A Survey of Multilingual Reasoning in Language Models](https://mlsanity.com/paper/2502.09457/) (2025)
- [mHumanEval -- A Multilingual Benchmark to Evaluate Large Language Models for Code Generation](https://mlsanity.com/paper/2410.15037/) (2024)
- [Revisiting Weight Averaging for Model Merging](https://mlsanity.com/paper/2412.12153/) (2024)
- [ProverbEval: Exploring LLM Evaluation Challenges for Low-resource Language Understanding](https://mlsanity.com/paper/2411.05049/) (2024)
- [Systematically Analyzing Prompt Injection Vulnerabilities in Diverse LLM Architectures](https://mlsanity.com/paper/2410.23308/) (2024)
- [Exploring Vulnerabilities and Protections in Large Language Models: A Survey](https://mlsanity.com/paper/2406.00240/) (2024)
- [On the Consistency of Multilingual Context Utilization in Retrieval-Augmented Generation](https://mlsanity.com/paper/2504.00597/) (2025)
- [Rethinking Weight-Averaged Model-merging](https://mlsanity.com/paper/2411.09263/) (2024)
- [Comparative Analysis of Multilingual Text Classification & Identification through Deep Learning and Embedding Visualization](https://mlsanity.com/paper/2312.03789/) (2023)

## Built on by (most cited first, 14 of 14)

- [Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought](https://mlsanity.com/paper/2510.04230/) (2025)
- [Evaluating Robustness of Large Language Models Against Multilingual Typographical Errors](https://mlsanity.com/paper/2510.09536/) (2025)
- [Beyond English-Centric Training: How Reinforcement Learning Improves Cross-Lingual Reasoning in LLMs](https://mlsanity.com/paper/2509.23657/) (2025)
- [R3S: Refining and Recovering Reinforcement Signals for Multilingual Understanding and Reasoning](https://mlsanity.com/paper/2602.05940/) (2026)
- [Parallel Scaling Law: Unveiling Reasoning Generalization through A Cross-Linguistic Perspective](https://mlsanity.com/paper/2510.02272/) (2025)
- [Language of Thought Shapes Output Diversity in Large Language Models](https://mlsanity.com/paper/2601.11227/) (2026)
- [The Reasoning Lingua Franca: A Double-Edged Sword for Multilingual AI](https://mlsanity.com/paper/2510.20647/) (2025)
- [Rethinking the Multilingual Reasoning Gap with Layer Swap](https://mlsanity.com/paper/2605.26735/) (2026)
- [The Fragility of Chain-of-Thought Monitoring Across Typologically Diverse Languages](https://mlsanity.com/paper/2605.27901/) (2026)
- [Reveal-Bangla: A Dataset for Cross-Lingual Multi-Step Reasoning Evaluation](https://mlsanity.com/paper/2508.08933/) (2025)
- [mmPISA-bench: Do LLMs Reason Equally Well Across 43 Languages?](https://mlsanity.com/paper/2606.07069/) (2026)
- [When Does Reasoning Help in Machine Translation? A Hierarchical Analysis of LRM Reasoning Traces](https://mlsanity.com/paper/2609.21247/) (2026)
- [What Does Post-Training Change in Multilingual Reasoning?](https://mlsanity.com/paper/2609.37104/) (2026)
- [Investigating the Role of Reasoning-Language Alignment in Monolingual Retrieval-Augmented Generation](https://mlsanity.com/paper/2610.03136/) (2026)

## Abstract (arXiv)

Recent Large Reasoning Models (LRMs) with thinking traces have shown strong performance on English reasoning tasks. However, their ability to think in other languages is less studied. This capability is as important as answer accuracy for real world applications because users may find the reasoning trace useful for oversight only when it is expressed in their own language. We comprehensively evaluate two leading families of LRMs on our XReasoning benchmark and find that even the most advanced models often revert to English or produce fragmented reasoning in other languages, revealing a substantial gap in multilingual reasoning. Prompt based interventions that force models to reason in the users language improve readability and oversight but reduce answer accuracy, exposing an important trade off. We further show that targeted post training on just 100 examples mitigates this mismatch, though some accuracy loss remains. Our results highlight the limited multilingual reasoning capabilities of current LRMs and outline directions for future work. Code and data are available at https://github.com/Betswish/mCoT-XReasoning.