arXiv cs.AI· Dalton Raphael Harmsen, Swier Garst, Thomas van Osch, Zar\`e Palanciyan, Joaquin Vanschoren·· 5 小时前AI 评分35
利用类型学特征代理实现零计算成本的跨语言迁移性估计
Zero-Compute Cross-Lingual Transferability Estimation Using Typological Feature Proxies
AI 导读
一项研究提出仅用类型学数据库特征通过随机森林预测跨语言迁移，无需大规模多语言预训练。在 24 语言迁移矩阵上，仅类型学模型留一语言交叉验证达 ρ=0.705、R²=0.49，与调优的非类型学对照（ρ=0.62）差异在配对 bootstrap 检验下不可区分；对未见语言对降至 ρ≈0.41。该方法以一次模型拟合替代数百次训练运行，可作零成本筛选工具，代码已开源。
正文 · 原文
Abstract:Cross-lingual transfer describes how knowledge in a source language benefits a target language. Measuring it quantitatively requires broad multilingual pre-training, as prior work has done with cross-lingual transfer matrices. We ask whether transfer is predictable from freely available typological features, and whether the prominence of high-resource source languages reflects typology or data quality and quantity. We show that typological databases contain cheap and dense signals about cross-lingual transfer. Our typology-only random forest on a 24-language prior-work transfer matrix scores leave-one-language-out $\rho{=}0.705$ and $R^2{=}0.49$, against $\rho{=}0.62$ for a tuned non-typological control, an increment that a paired bootstrap cannot separate at this scale. When both languages of a held-out pair are unseen, performance degrades to $\rho{\approx}0.41$, so the model interpolates a ground-truth matrix more reliably than it predicts new languages. The signal survives leave-one-script-out and leave-one-family-out protocols, so script and family confounding do not explain the effect. By decomposing the transfer into a typology term and a resource-and-script bias term, we find the best-source ranking sensitive to this bias. Typological features never observe this bias, which makes them a zero-compute screening tool that replaces hundreds of training runs with a model fit. Our code is available \href{this https URL}{here}.
| Comments: | 4 pages, NeurIPS workshop, Linguistic Principles for Foundation Models, lp4fm |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2609.39640 [cs.CL] |
| (or arXiv:2609.39640v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2609.39640 arXiv-issued DOI via DataCite |
Submission history
From: Dalton Harmsen [view email]
[v1]
Wed, 30 Sep 2026 12:45:46 UTC (9,166 KB)
[v2]
Wed, 4 Nov 2026 13:00:52 UTC (9,167 KB)
来源：arXiv cs.AI · arxiv.org