Matches in SemOpenAlex for { <https://semopenalex.org/work/W4378464930> ?p ?o ?g. }
Showing items 1 to 65 of
65
with 100 items per page.
- W4378464930 abstract "Recent multilingual pretrained language models (mPLMs) have been shown to encode strong language-specific signals, which are not explicitly provided during pretraining. It remains an open question whether it is feasible to employ mPLMs to measure language similarity, and subsequently use the similarity results to select source languages for boosting cross-lingual transfer. To investigate this, we propose mPLM-Sim, a new language similarity measure that induces the similarities across languages from mPLMs using multi-parallel corpora. Our study shows that mPLM-Sim exhibits moderately high correlations with linguistic similarity measures, such as lexicostatistics, genealogical language family, and geographical sprachbund. We also conduct a case study on languages with low correlation and observe that mPLM-Sim yields more accurate similarity results. Additionally, we find that similarity results vary across different mPLMs and different layers within an mPLM. We further investigate whether mPLM-Sim is effective for zero-shot cross-lingual transfer by conducting experiments on both low-level syntactic tasks and high-level semantic tasks. The experimental results demonstrate that mPLM-Sim is capable of selecting better source languages than linguistic measures, resulting in a 1%-2% improvement in zero-shot cross-lingual transfer performance." @default.
- W4378464930 created "2023-05-27" @default.
- W4378464930 creator A5014957657 @default.
- W4378464930 creator A5038139314 @default.
- W4378464930 creator A5049217765 @default.
- W4378464930 creator A5051693368 @default.
- W4378464930 creator A5071144367 @default.
- W4378464930 date "2023-05-23" @default.
- W4378464930 modified "2023-09-25" @default.
- W4378464930 title "mPLM-Sim: Unveiling Better Cross-Lingual Similarity and Transfer in Multilingual Pretrained Language Models" @default.
- W4378464930 doi "https://doi.org/10.48550/arxiv.2305.13684" @default.
- W4378464930 hasPublicationYear "2023" @default.
- W4378464930 type Work @default.
- W4378464930 citedByCount "0" @default.
- W4378464930 crossrefType "posted-content" @default.
- W4378464930 hasAuthorship W4378464930A5014957657 @default.
- W4378464930 hasAuthorship W4378464930A5038139314 @default.
- W4378464930 hasAuthorship W4378464930A5049217765 @default.
- W4378464930 hasAuthorship W4378464930A5051693368 @default.
- W4378464930 hasAuthorship W4378464930A5071144367 @default.
- W4378464930 hasBestOaLocation W43784649301 @default.
- W4378464930 hasConcept C103278499 @default.
- W4378464930 hasConcept C115961682 @default.
- W4378464930 hasConcept C130318100 @default.
- W4378464930 hasConcept C137293760 @default.
- W4378464930 hasConcept C138885662 @default.
- W4378464930 hasConcept C154945302 @default.
- W4378464930 hasConcept C173608175 @default.
- W4378464930 hasConcept C204321447 @default.
- W4378464930 hasConcept C2776175482 @default.
- W4378464930 hasConcept C2776517306 @default.
- W4378464930 hasConcept C2780813799 @default.
- W4378464930 hasConcept C41008148 @default.
- W4378464930 hasConcept C41895202 @default.
- W4378464930 hasConcept C46686674 @default.
- W4378464930 hasConceptScore W4378464930C103278499 @default.
- W4378464930 hasConceptScore W4378464930C115961682 @default.
- W4378464930 hasConceptScore W4378464930C130318100 @default.
- W4378464930 hasConceptScore W4378464930C137293760 @default.
- W4378464930 hasConceptScore W4378464930C138885662 @default.
- W4378464930 hasConceptScore W4378464930C154945302 @default.
- W4378464930 hasConceptScore W4378464930C173608175 @default.
- W4378464930 hasConceptScore W4378464930C204321447 @default.
- W4378464930 hasConceptScore W4378464930C2776175482 @default.
- W4378464930 hasConceptScore W4378464930C2776517306 @default.
- W4378464930 hasConceptScore W4378464930C2780813799 @default.
- W4378464930 hasConceptScore W4378464930C41008148 @default.
- W4378464930 hasConceptScore W4378464930C41895202 @default.
- W4378464930 hasConceptScore W4378464930C46686674 @default.
- W4378464930 hasLocation W43784649301 @default.
- W4378464930 hasOpenAccess W4378464930 @default.
- W4378464930 hasPrimaryLocation W43784649301 @default.
- W4378464930 hasRelatedWork W1567769573 @default.
- W4378464930 hasRelatedWork W1633328346 @default.
- W4378464930 hasRelatedWork W1997312918 @default.
- W4378464930 hasRelatedWork W2056556847 @default.
- W4378464930 hasRelatedWork W2102964821 @default.
- W4378464930 hasRelatedWork W2172323827 @default.
- W4378464930 hasRelatedWork W2293496125 @default.
- W4378464930 hasRelatedWork W2951665417 @default.
- W4378464930 hasRelatedWork W4281690070 @default.
- W4378464930 hasRelatedWork W4287890973 @default.
- W4378464930 isParatext "false" @default.
- W4378464930 isRetracted "false" @default.
- W4378464930 workType "article" @default.