Recent strategies for low-resource machine translation rely on LLMs to generate synthetic data from higher-resource languages. We find that this method fails for Romansh, because LLMs tend to confuse its 6 distinct language varieties. Our experiments show that instead, the direction of data augmentation should be aligned with the resource gradient between source and target language. This approach surpasses Gemini 3 Pro in the lowest-resource variety of Romansh by 23 BLEU. A human evaluation confirms that our experiments yield the first model that generates fluent translations in the individual Romansh varieties.

翻译：摘要：近期面向低资源机器翻译的策略依赖大语言模型从高资源语言生成合成数据。我们发现该方法在罗曼什语上失效，原因是LLM容易混淆其6种不同语言变体。实验表明，数据增强的方向应与源语言与目标语言之间的资源梯度对齐。该方案在罗曼什语最低资源变体上以23 BLEU值超越Gemini 3 Pro。人工评估证实，我们的实验生成了首个能生成各罗曼什语变体流畅翻译的模型。