Unverified50% confidenceFactExact time
For low-resource language translation, the most effective way to use monolingual data is not continued pre-training, but back-translation.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe advantage of continued pre-training is that it can exploit massive unlabeled monolingual data, but the cost is that it requires a large amount of compute, and what is learned is only the distribution at the level of language form, not directly targeted at translation, the cross-lingual mapping task.75% similarUnverifiedEnd-to-end speech translation requires large amounts of aligned data pairing source-language speech with target-language text, which is extremely scarce for low-resource language pairs.75% similarUnverifiedA parallel corpus is a bilingual aligned dataset composed of source-language sentences and their corresponding target-language translations, and is the core training material for machine translation.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1028052API
curl https://kongchang.com/api/v1/knowledge/claims/1028052MCP
get_claim(id=1028052)