← 研踪 PaperTrailarXiv:2606.21954 · v2
Are Multilingual Models Actually Improving? Revisiting Cross-Lingual Transfer
首次提交 2026-06-20 17:03
版本更新 2026-11-09 14:44
雷达观测 2026-11-10 13:18
摘要
Cross-lingual transfer is a model's ability to generalize capabilities from well-represented source languages to under-represented target languages. Existing measures of a model's transfer strength conflate improvements in transfer with general improvements to accuracy in the source language. We advocate for an alternate metric that reliably captures transfer strength called Hardness Adjusted Transfer (HAT) Score, and use it to derive multiple insights on factors influencing transfer strength. Our analysis across twenty diverse language models and three popular mainstream multilingual benchmarks argues that 1) transfer in small models is not broken, 2) we are making slower than expected progress in cross-lingual transfer with model size, and 3) we have made clear progress over time.
期刊信息：https://aclanthology.org/2026.findings-emnlp.2388/
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v2 · 2026-11-09 14:44 · 最新观测版本
Are Multilingual Models Actually Improving? Revisiting Cross-Lingual Transfer
Cross-lingual transfer is a model's ability to generalize capabilities from well-represented source languages to under-represented target languages. Existing measures of a model's transfer strength conflate improvements in transfer with general improvements to accuracy in the source language. We advocate for an alternate metric that reliably captures transfer strength called Hardness Adjusted Transfer (HAT) Score, and use it to derive multiple insights on factors influencing transfer strength. Our analysis across twenty diverse language models and three popular mainstream multilingual benchmarks argues that 1) transfer in small models is not broken, 2) we are making slower than expected progress in cross-lingual transfer with model size, and 3) we have made clear progress over time.
查看此版本解读 →查看此版本原文 ↗