# ReasonXL: Shifting LLM Reasoning Language Without Sacrificing Performance

arXiv: https://arxiv.org/abs/2604.12378 · HTML: https://mlsanity.com/paper/2604.12378/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 16 of its 62 references that are in the map.
- First posted on arXiv 14 Apr 2026.

## Placed in

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

## Builds on (most cited first, 16 of 16)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [Llama-Nemotron: Efficient Reasoning Models](https://mlsanity.com/paper/2505.00949/) (2025)
- [Aya Expanse: Combining Research Breakthroughs for a New Multilingual Frontier](https://mlsanity.com/paper/2412.04261/) (2024)
- [Olmo 3](https://mlsanity.com/paper/2512.13961/) (2025)
- [NVIDIA Nemotron 3: Efficient and Open Intelligence](https://mlsanity.com/paper/2512.20856/) (2025)
- [Magistral](https://mlsanity.com/paper/2506.10910/) (2025)
- [Nemotron-Cascade: Scaling Cascaded Reinforcement Learning for General-Purpose Reasoning Models](https://mlsanity.com/paper/2512.13607/) (2025)
- [Crosslingual Reasoning through Test-Time Scaling](https://mlsanity.com/paper/2505.05408/) (2025)
- [Language Matters: How Do Multilingual Input and Reasoning Paths Affect Large Reasoning Models?](https://mlsanity.com/paper/2505.17407/) (2025)
- [Distribution-Aligned Sequence Distillation for Superior Long-CoT Reasoning](https://mlsanity.com/paper/2601.09088/) (2026)
- [Language Models' Factuality Depends on the Language of Inquiry](https://mlsanity.com/paper/2502.17955/) (2025)
- [CLaS-Bench: A Cross-Lingual Alignment and Steering Benchmark](https://mlsanity.com/paper/2601.08331/) (2026)
- [Gained in Translation: Privileged Pairwise Judges Enhance Multilingual Reasoning](https://mlsanity.com/paper/2601.18722/) (2026)
- [propella-1: Multi-Property Document Annotation for LLM Data Curation at Scale](https://mlsanity.com/paper/2602.12414/) (2026)

## Built on by (most cited first, 5 of 5)

- [Rethinking the Multilingual Reasoning Gap with Layer Swap](https://mlsanity.com/paper/2605.26735/) (2026)
- [Cost of Reasoning in non-English Languages: A Case Study on Japanese](https://mlsanity.com/paper/2607.10114/) (2026)
- [AdaMame: A Training Recipe for Adaptive Multilingual Reasoning](https://mlsanity.com/paper/2606.15080/) (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)

## Abstract (arXiv)

Despite advances in multilingual capabilities, most large language models (LLMs) remain English-centric in their training and, crucially, in the reasoning traces they produce: even for non-English problems, they predominantly reason in English, leaving non-English users with intermediate reasoning they cannot follow. We address this disparity with two contributions. **(a)** We introduce textttReasonXL, the first large-scale parallel corpus of cross-domain reasoning traces in five European languages (English, German, French, Italian, and Spanish): over 2.5M aligned samples (11B tokens) per language, each comprising prompt, reasoning trace, and final output, enabling direct supervision of language-specific reasoning. **(b)** Using textttReasonXL, we demonstrate that LLMs can be adapted to reason entirely in a target language through a simple two-stage pipeline of supervised fine-tuning (SFT) and reinforcement learning with verifiable rewards (RLVR), matching or exceeding the base models' English-reasoning performance in two of four languages and recovering most of it in the remaining two.