# Think Natively: Unlocking Multilingual Reasoning with Consistency-Enhanced Reinforcement Learning

arXiv: https://arxiv.org/abs/2510.07300 · HTML: https://mlsanity.com/paper/2510.07300/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 29 of its 32 references that are in the map.
- First posted on arXiv 8 Oct 2025.

## Placed in

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

## Builds on (most cited first, 29 of 29)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [Let's Verify Step by Step](https://mlsanity.com/paper/2305.20050/) (2023)
- [HybridFlow: A Flexible and Efficient RLHF Framework](https://mlsanity.com/paper/2409.19256/) (2024)
- [LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models](https://mlsanity.com/paper/2403.13372/) (2024)
- [A Survey on LLM-as-a-Judge](https://mlsanity.com/paper/2411.15594/) (2024)
- [DeepSeek-V3 Technical Report](https://mlsanity.com/paper/2412.19437/) (2024)
- [Statistical Rejection Sampling Improves Preference Optimization](https://mlsanity.com/paper/2309.06657/) (2023)
- [Light-R1: Curriculum SFT, DPO and RL for Long COT from Scratch and Beyond](https://mlsanity.com/paper/2503.10460/) (2025)
- [Competitive Programming with Large Reasoning Models](https://mlsanity.com/paper/2502.06807/) (2025)
- [MAPO: Advancing Multilingual Reasoning through Multilingual Alignment-as-Preference Optimization](https://mlsanity.com/paper/2401.06838/) (2024)
- [Magistral](https://mlsanity.com/paper/2506.10910/) (2025)
- [PolyMath: Evaluating Mathematical Reasoning in Multilingual Contexts](https://mlsanity.com/paper/2504.18428/) (2025)
- [A Survey on Large Language Models with Multilingualism: Recent Advances and New Frontiers](https://mlsanity.com/paper/2405.10936/) (2024)
- [Language Imbalance Driven Rewarding for Multilingual Self-improving](https://mlsanity.com/paper/2410.08964/) (2024)
- [Learn Globally, Speak Locally: Bridging the Gaps in Multilingual Reasoning](https://mlsanity.com/paper/2507.05418/) (2025)
- [Language Mixing in Reasoning Language Models: Patterns, Impact, and Internal Causes](https://mlsanity.com/paper/2505.14815/) (2025)
- [Language Matters: How Do Multilingual Input and Reasoning Paths Affect Large Reasoning Models?](https://mlsanity.com/paper/2505.17407/) (2025)
- [Multilingual Knowledge Editing with Language-Agnostic Factual Neurons](https://mlsanity.com/paper/2406.16416/) (2024)
- [AlignDistil: Token-Level Language Model Alignment as Adaptive Policy Distillation](https://mlsanity.com/paper/2503.02832/) (2025)
- [MMATH: A Multilingual Benchmark for Mathematical Reasoning](https://mlsanity.com/paper/2505.19126/) (2025)
- [ExTrans: Multilingual Deep Reasoning Translation via Exemplar-Enhanced Reinforcement Learning](https://mlsanity.com/paper/2505.12996/) (2025)
- [Cross-lingual Collapse: How Language-Centric Foundation Models Shape Reasoning in Large Language Models](https://mlsanity.com/paper/2506.05850/) (2025)
- [DeepTrans: Deep Reasoning Translation via Reinforcement Learning](https://mlsanity.com/paper/2504.10187/) (2025)
- [A Dual-Space Framework for General Knowledge Distillation of Large Language Models](https://mlsanity.com/paper/2504.11426/) (2025)
- [Making Qwen3 Think in Korean with Reinforcement Learning](https://mlsanity.com/paper/2508.10355/) (2025)
- [CM-Align: Consistency-based Multilingual Alignment for Large Language Models](https://mlsanity.com/paper/2509.08541/) (2025)

## Built on by (most cited first, 18 of 18)

- [Crosslingual On-Policy Self-Distillation for Multilingual Reasoning](https://mlsanity.com/paper/2605.09548/) (2026)
- [TAPO: Translation Augmented Policy Optimization for Multilingual Mathematical Reasoning](https://mlsanity.com/paper/2603.25419/) (2026)
- [What Makes Good Multilingual Reasoning? Disentangling Traces with Measurable Features](https://mlsanity.com/paper/2604.04720/) (2026)
- [LANG: Reinforcement Learning for Multilingual Reasoning with Language-Adaptive Hint Guidance](https://mlsanity.com/paper/2605.22567/) (2026)
- [Language-Coupled Reinforcement Learning for Multilingual Retrieval-Augmented Generation](https://mlsanity.com/paper/2601.14896/) (2026)
- [Rethinking the Multilingual Reasoning Gap with Layer Swap](https://mlsanity.com/paper/2605.26735/) (2026)
- [Efficient Multilingual Reasoning Transfer via Progressive Code-Switching](https://mlsanity.com/paper/2607.00485/) (2026)
- [Learning to Route Languages for Multilingual Policy Optimization](https://mlsanity.com/paper/2605.25360/) (2026)
- [PolyFact: Comparing Consistency-Driven Post-training Methods for Cross-Lingual Factual Recall](https://mlsanity.com/paper/2606.06586/) (2026)
- [Beyond Input Understanding: Diagnosing Multilingual Mathematical Reasoning with Directed Acyclic Trace Graphs](https://mlsanity.com/paper/2605.27715/) (2026)
- [Translate-R1: Cost-Aware Translation Tool Use via Reinforcement Learning](https://mlsanity.com/paper/2606.06835/) (2026)
- [Soft Token Alignment for Cross-Lingual Reasoning](https://mlsanity.com/paper/2606.26466/) (2026)
- [Distill Where the Student Goes: Teacher-Regularized RL for English-Evidence Cross-Lingual RAG](https://mlsanity.com/paper/2607.02966/) (2026)
- [GRPO Beyond English: A Large-Scale Study of GRPO in Non-English and Multilingual Settings](https://mlsanity.com/paper/2608.13698/) (2026)
- [Thinking in a Low-Resource Language: What SFT Builds, What RL Fixes, What Accuracy Cannot See](https://mlsanity.com/paper/2608.17744/) (2026)
- [Building Multilingual Bridges: Data Mixing as the Pillar of Generalization for In-Language Reasoning](https://mlsanity.com/paper/2609.10445/) (2026)
- [What Does Post-Training Change in Multilingual Reasoning?](https://mlsanity.com/paper/2609.37104/) (2026)
- [MixLID: Token-Level Language Identification in Mixed-Language Text](https://mlsanity.com/paper/2610.16601/) (2026)

## Abstract (arXiv)

Large Reasoning Models (LRMs) have achieved remarkable performance on complex reasoning tasks by adopting the ``think-then-answer'' paradigm, which enhances both accuracy and interpretability. However, current LRMs exhibit two critical limitations when processing non-English languages: (1) They often struggle to maintain input-output language consistency; (2) They generally perform poorly with wrong reasoning paths and lower answer accuracy compared to English. These limitations significantly compromise the interpretability of reasoning processes and degrade the user experience for non-English speakers, hindering the global deployment of LRMs. To address these limitations, we propose M-Thinker, which is trained by the GRPO algorithm that involves a Language Consistency (LC) reward and a novel Cross-lingual Thinking Alignment (CTA) reward. Specifically, the LC reward defines a strict constraint on the language consistency between the input, thought, and answer. Besides, the CTA reward compares the model's non-English reasoning paths with its English reasoning path to transfer its own reasoning capability from English to non-English languages. Through an iterative RL procedure, our M-Thinker-1.5B/4B/7B models not only achieve nearly 100% language consistency and superior performance on two multilingual benchmarks (MMATH and PolyMath), but also exhibit excellent generalization on out-of-domain languages.