SOTAVerified

Cross-Lingual Transfer

Cross-lingual transfer refers to transfer learning using data and models available for one language for which ample such resources are available (e.g., English) to solve tasks in another, commonly more low-resource, language.

Papers

Showing 221230 of 782 papers

TitleStatusHype
Beyond the English Web: Zero-Shot Cross-Lingual and Lightweight Monolingual Classification of RegistersCode0
Multilingual Content Moderation: A Case Study on RedditCode0
Cross-lingual Offensive Language Detection: A Systematic Review of Datasets, Transfer Approaches and ChallengesCode0
A Simple and Effective Method To Eliminate the Self Language Bias in Multilingual RepresentationsCode0
Contextual Label Projection for Cross-Lingual Structured PredictionCode0
Massively Multilingual Transfer for NERCode0
A Simple and Effective Method to Improve Zero-Shot Cross-Lingual Transfer LearningCode0
Exploring Multilingual Concepts of Human Value in Large Language Models: Is Value Alignment Consistent, Transferable and Controllable across Languages?Code0
Detecting and Mitigating Hallucinations in Multilingual SummarisationCode0
Analysing The Impact Of Linguistic Features On Cross-Lingual TransferCode0
Show:102550
← PrevPage 23 of 79Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1PaLM 2 (few-shot)Accuracy94.4Unverified
2mT0-13BAccuracy84.45Unverified
3RoBERTa Large (translate test)Accuracy76.05Unverified
4BLOOMZAccuracy75.5Unverified
5MAD-X BaseAccuracy60.94Unverified
6mGPTAccuracy55.5Unverified