--- title: "Analyzing and Improving Cross-lingual Knowledge — Stap 2026" description: "Most of the world's languages have little parallel data, so translating them depends on multilingual models reusing knowledge learned from…" authors: ["David Stap"] year: 2026 venue: "Proceedings of the 26th Annual Conference of the European Association for Machine Translation (Volume 1)" source: "https://aclanthology.org/2026.eamt-1.2/" licence: "CC BY-ND 4.0 (EAMT)" licence_url: "https://creativecommons.org/licenses/by-nd/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/stap-2026-analyzing-improving-cross-lingual-knowledge/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/stap-2026-analyzing-improving-cross-lingual-knowledge/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/) / Analyzing and Improving Cross-lingual Knowledge Transfer for Machine Translation # Analyzing and Improving Cross-lingual Knowledge Transfer for Machine Translation **Authors:** David Stap · **Year:** 2026 · **Venue:** Proceedings of the 26th Annual Conference of the European Association for Machine Translation (Volume 1) · **Licence:** [CC BY-ND 4.0 (EAMT)](https://creativecommons.org/licenses/by-nd/4.0/) **Links:** [Publisher's page](https://aclanthology.org/2026.eamt-1.2/) · [PDF](https://aclanthology.org/2026.eamt-1.2.pdf) · [OpenAlex](https://openalex.org/W7119531192) **Citations:** 0 citations (openalex-snapshot · 2026-10-14) ## Summary Most of the world's languages have little parallel data, so translating them depends on multilingual models reusing knowledge learned from better-resourced languages. This EAMT 2026 thesis abstract by David Stap (University of Amsterdam) studies that cross-lingual transfer from four angles. 1. It introduces Representational Transfer Potential (RTP), a measure based on how similar cross-attention context vectors are in multilingual NMT. It then adds an auxiliary loss that uses multi-parallel data to bring language representations closer together. 2. It adapts k-NN-MT to cross-lingual and multilingual datastores, using linear mappings to align representations across languages. 3. It examines how fine-tuning LLMs (7B–65B) on parallel data erodes skills such as formality control, and counters this by adding monolingual text to the training mix. 4. It varies language diversity across 132 translation directions and evaluates with BLEU and COMET. The findings indicate that transfer happens at the representational level. Broader language coverage also helps unseen pairs, and mixed-data training preserves LLM abilities while improving quality. Together, these results give practical guidance for building translation for under-resourced languages. ## Key points - RTP correlates strongly with translation gains (Spearman's ρ = .77, p \< 0.001). Multi-parallel overlap emerges as an under-explored predictor of transfer, alongside subword overlap and genetic distance. - In k-NN-MT, datastores built from related high-resource languages beat bilingual ones for low-resource pairs. Language-group datastores match full multilingual ones while being up to 75% smaller and up to 5.3× faster. - Parallel-only LLM fine-tuning harms formality steering, few-shot domain adaptation and document-level translation. The damage appears with as few as 18K examples and at every model scale tested. - Mixed-data fine-tuning, combining parallel and monolingual text, keeps these capabilities and yields higher overall translation quality. - Greater language diversity during fine-tuning improves both seen and unseen pairs and increases cross-lingual overlap in middle layers. The author releases the IdiomsInCtx-MT dataset, code and model checkpoints. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* Fields: [Linguistics & languages](https://md.research.elamur.ai/free/en/shelves/linguistics/) · [Machine translation](https://md.research.elamur.ai/free/en/shelves/machine-translation/) Topics: [Machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/) ## Cite The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{stap2026analyzing, title = {{Analyzing and Improving Cross-lingual Knowledge Transfer for Machine Translation}}, author = {Stap, David}, booktitle = {Proceedings of the 26th Annual Conference of the European Association for Machine Translation (Volume 1)}, year = {2026}, url = {https://aclanthology.org/2026.eamt-1.2/} } ``` - [← previous An LLM-Based Assistant for Debt Waiver Court Procedures](https://md.research.elamur.ai/free/en/papers/padro-2026-llm-based-assistant-debt-waiver/) - [next → Anchoring the Judge: Curriculum-Based Adaptation and Reference-Anchored MQM for LLM-Based Machine Translation of an Unseen Low-Resource Language - A Case of Nupe](https://md.research.elamur.ai/free/en/papers/umar-2026-anchoring-judge-curriculum-based-adaptation/) - [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/) ## References 1 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. [Stap and Monz2023] Stap, David and Christof Monz. 2023. Multilingual k-nearest-neighbor machine translation. In Bouamor, Houda, Juan Pino, and Kalika Bali, editors, Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 9200–9208, Singapore, December. Associat… [find](https://scholar.google.com/scholar?q=%5BStap+and+Monz2023%5D+Stap%2C+David+and+Christof+Monz%2E+2023%2E+Multilingual+k%2Dnearest%2Dneighbor+machine+translation%2E+In+Bouamor%2C+Houda%2C+Juan+Pino%2C+and+Kalika+Bali%2C+editors%2C+Proceedings+of+the+2023+Conference)