Verified75% confidenceFactTime unknown
In 2016, Google Translate fully transitioned from Statistical Machine Translation (SMT) to deep neural networks.
3
Sources
75%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedGoogle's Neural Machine Translation system (GNMT) is based on a multi-layer LSTM plus attention architecture, and RNN-based models dominated the field of machine translation between 2014 and 201670% similarUnverified谷歌翻译在2016年大规模切换至LSTM架构后翻译质量大幅提升66% similarUnverified2022年以来Freitag等人在Google的研究表明，基于神经效用指标（如BLEURT、COMET）的MBR解码在WMT翻译任务上持续超越束搜索65% similarUnverifiedGoogle Cloud Translation API是谷歌基于神经机器翻译（NMT）技术构建的云端翻译服务，支持100余种语言65% similarUnverifiedAfter Google introduced the Transformer architecture in 2017, the quality of neural machine translation (NMT) improved dramatically.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/22943API
curl https://kongchang.com/api/v1/knowledge/claims/22943MCP
get_claim(id=22943)